Drug Testing

TRANSCRIPT: ON-SITE DRUG TESTING WORKGROUP MEETING (October 5-6, 1999)
Substance Abuse and Mental Health Services Administration

AGENDA (October 5)

• Welcome and Introduction - Dr. Donna Bush
• Factors for Accurate and Reliable Workplace Drug Testing - Dr. Yale Caplan
  
• Studies on Performance of Non-Instrumented Drug Test Devices - Dr. Robert Willette
  
• Two Readers of NIDT Results - Concordance and Accuracy - Dr. Leo Kadehjian
  
• Parole and Probation Experiences and Model Program - Mr. Jim Baer
  
• District of Columbia Pre-Trial, Probation and Drug Court Experience - Mr. Jay Carver
  
• On-Site Drug Testing Program Oversight - Dr. Leo
  
• Kadehjian and Dr. Robert Willette
  
• Questions and Answers
  
• Manufacturers and Distributors Presentations
  

DR. BUSH (HHS): Welcome. We are here today as a follow on to the Drug Testing Advisory Board. The Drug Testing Advisory Board for the last two and a half years has been examining alternative specimens and technologies for the purposes of examining illegal drug use and abuse in the workplace. Today, we are taking a look at on-site testing technologies as they apply to both urine and oral fluid. It is going to be possibly laboratory, on-site laboratory based, and possibly just using non-instrumented devices. These are the concepts that are floating out there that have been presented to the Drug Testing Advisory Board for their review. Because of the sheer number of people who have products and ideas and experience with these testing technologies and implementing it and using it in the workplace, we wanted to gather a large audience to help us gather as much information as we can. It is a working group of the Drug Testing Advisory Board.

I am starting the meeting today, but Dr. Yale Caplan and Dr. Bob Willette will be the co-chairs. Dr. Caplan will chair today's events and Dr. Willette will chair tomorrow's.

You have some handouts that you picked up as you registered. Please, register with us. That will help us forward information to you in the future.

We have some requests, official requests, from the chair. It has to do with how you participate with us, if you are going to ask a question, come to the microphone. Wait to be recognized. State your name and affiliation for the record, and please make sure we have your business address.

This entire proceeding will be transcribed and placed on our Internet Web site. Currently, that resident Web site address is WWW.HEALTH.ORG/WORKPL.HTM. It is going to take us about a month to get the original transcription and then several days for us to make it readable without changing the content.

FACTORS FOR ACCURATE AND RELIABLE WORKPLACE DRUG TESTING

DR. CAPLAN (Board Member): A few notes of historical perspective. As you were told, Bob and I were asked to chair this group and this working group. As you all know, if you have attended before, this meeting is one of a continuum of working groups. The idea was to try to formulate various ways to get the right people together to get the input so that ultimately, the information could be transcribed to the DTAB for further consideration.

As you have seen with some of the other groups where the number of participants was relatively low, there were more off site working groups just of participants, but due to the large number here, we came back to a format which was similar to the format that we used about two and a half years ago for the initial meetings. The idea is to get some general overviews from a group of people who have practical experiences and then an opportunity for manufacturers and providers of either products or services to tell us about how they are doing it.

Historically, two and a half years ago another working group led by Dave Evans, which is much more of an industry based group, the NOTA Group, will give us a brief update this afternoon. However, the nature of on-site testing being originally focused principally on urine and there not being quite as many fundamental questions about cutoffs and the specimen itself, that group didn't have a need or do a great deal over the relevant early periods of time, but now, we are going to bring it back together in this kind of format.

Another thing I'd like to bring to some people's attention is the 1989 NIDA Consensus Conference, a thought long to be forgotten, but I think it's really back. I would like to read a few things that were said then that pertain to on-site drug testing, and find, I think, that we kind of said the same things then that we are saying today. I think what will be appreciated is that we are probably going to bring them to some conclusion after about ten years.

In the introduction there, there was a statement that as authorized by the Executive Order 12564, et cetera, mandatory guidelines were established, establishing Federal standards for conducting urine drug testing for Federal employees and certification of standards for laboratories to test these specimens. The institution of these standards, together with the NLCP, National Laboratory Certification Program, was novel at the time for clinical chemists and toxicologists. Again, I think there are certain novel things now that we are moving forward into these alternative procedures. It established a precedent and sound basis requiring good laboratory practices in conducting employee drug testing, in cases which must withstand legal scrutiny.

It was difficult and stressful for those laboratory scientists involved at the outset. Well, nothing has changed. It continues to be difficult and stressful to go through this revised process.

At that time, scientists from NIDA, together with forensic toxicologists, worked steadily to find a practical laboratory program, expanding, constricting, defining, refining, and finally culminating in a paradigm which permits the testing of human urine for five commonly abused drugs, with a minimum of error and a maximum of protection for individual employees. No single event, individual, discovery or discipline can be identified as a sole contributor to this accomplishment. Rather, a combination of technological advances and management techniques were applied in a contents of forensic science to a rapidly expanding drug abuse program of urgent concern to society. Well, nothing has changed. We are essentially doing the same thing again, but broadening the horizon.

There were some general recommendations at that time in 1989, ten years ago. A few of these are worth putting into the record.

On-site initial screening only testing facilities. The facility should only be allowed where safety issues demand the most rapid turnaround time, justifying the risk to the client inherent in unconfirmed tests, and the considerable difficulties in achieving accurate testing that such facilities create.

Secondly, on-site urine screening can reliably identify negative specimens, provided appropriate safeguards are built into the procedure. These precautions include meeting the basic forensic standards for specimen collection, chain of custody, documentation and security, splitting the collected urine specimen into two portions, participation in open and blind proficiency testing, a rigorous quality assurance program, being subject to on-site inspections, using an FDA approved screening test that provides objective and documentable results, use of the same cutoff concentrations as used in NIDA, now HHS SAMHSA certified laboratories, submitting all presumptive positive specimens to a NIDA certified laboratory for confirmation.

The last of these was if laboratory testing is performed on-site, all MRO functions should remain the same as at present, and other recommendations are given elsewhere.

I think we are here today to deal with some of these "other recommendations," that quite frankly were there and made ten years ago, but have finally come to fruition. In order to facilitate that process, we have revised the grid. As you know, we have been working with a kind of a grid dealing with the factors associated with workplace testing. There are some elements of that grid which applied mostly to the prior specimens and materials that we dealt with. However, for the on-site thing, a couple of things happened. Number one, it became apparent that on-site would include oral fluids. There will be an oral fluid element on-site, which probably wasn't anticipated initially. Also, on-site testing appears to deal with two types, the non- instrumented based techniques and possibly, instrumented based techniques. Therefore, for the purposes of the Board's ultimate evaluation, we created a revised grid, including some of the older elements, but certainly leaving things out like the lab certification and elements that are clearly laboratory based, and provided you with a new grid. The grid also has had the i's and p's and blanks temporarily removed. You probably will recall in the prior grid, there were a number of places where these letters appeared. Again, it was decided we will take a fresh start with looking at these comprehensively across the grid and fill it in again. Many of the things that are on this grid have already been discussed. A number of them have not. They kind of follow the pattern that we were asked to do many, many years ago and are finally doing now.

I'm just going to take a couple of minutes to go through the grid, because we are not ready to talk about it. This grid will be used by the Board at a future meeting when the content of this meeting's presentations from both the general speakers and the industry specific speakers are available. There will be some assimilation of this, and this is the grid we will start to ask the same questions about and see whether the information is there. That is the format the Board will use at later meetings that are focused on this topic.

To go through them quickly, the things that were in there -- part of the reasons we are having it this way is that in the original grid, there was only on-site urine. Now that we are thinking about on-site associated with oral fluid and possibly instrumented, the questions resurfaced themselves, so we created a column for both on-site instrumented and non-instrumented oral fluid and on-site instrumented and non-instrumented urine, dealing with a number of the same issues we had before, and the first few pages are fairly redundant and I'm not going to spend a lot of time on them. They deal with things that we know and have looked at already.

Preparation of the collection site, the security, privacy, and ability to do an observed collection. These may be more relevant as observed collection becomes more of a comparative issue with oral fluid compared to urine. Issues about the collector and the collector is later defined as perhaps the technician or tester, and the level of their training or certification.

The next overhead shows the device. We will again be looking at the container. In looking at oral fluid and the information from that early working group, there are a number of types of containers and a number of container related issues dealing with security, the FDA's clearance of those products, and whether or not those devices adversely affect or appropriately maintain the integrity of the specimen.

The specimen itself, in the case of urine, it is no different than the basic lab specimen, but in the case of oral fluid there are some issues which we will integrate from the other working group, the site and nature of the specimen's collection, whether accurate volume can be there to permit multiple and repeat testing, whether specimens can be effectively split and what kind of devices are needed for that. This is an area question which is probably a more global concern, that we haven't discussed before.

If we are going to do on-site testing, exactly how is this specimen going to be split and made available for future testing, and the questions of stability and storage, both on-site prior to testing and during the process.

The collection procedure itself includes the traditional elements of donor identification and ensuring that, and preparation of the donor for the specimen collection. Paramount here will be the issue of the donor's identity and the knowledge of the donor to the actual testing and how that all works.

We have the other elements of the procedure, chain of custody, the appropriate development of documentation, the custody form, and the major issue of specimen integrity evaluation.

As we move forward with the concerns we have with adulteration/specimen integrity today in the basic program, that will be a major place to see where that can be dealt with more effectively or at least as effectively by the use of these on-site devices. The transportation, deterring tampering, and adulteration.

Then we have the section designated more for the differences in on-site testing. The others were common from before. Do we want these products to be FDA cleared? What does that mean? There has been a lot of activity at FDA regarding these types of products in recent times.

The drug class specific. Whether the specificity of the drug is the same, although again, we may be using similar immunoassays. Their specificity will be an issue.

Documentation of collection will be discussed.

In the next area, further information about how to document the results, whether these procedures can be done, whether they are going to be done directly on the site, the nature of the form, whether they are going to be done electronically, and whether or not, finally, the HHS target analytes of concern are going to be the same and certainly with oral fluid, that may be an issue. With urine, it may not.

Probably a major question in this is objective differentiation of positive and negative. This is an area of concern in that clearly laboratories have been asked to be 25 percent above and 25 percent below cutoffs and then that question has to be dealt with and transcribed into what is the ability of an on-site product when compared to a laboratory to do that.

Another major issue will be the training and certification both of sites and of analysts at sites. This is where we do need a lot of information. If we are moving the focus from the laboratory, which is highly regulated, to an on-site area, that is going to be a major question area for the Board.

As I mentioned earlier, some of these questions will eventually be administrative questions, but the donor identity question needs to be dealt with, and again, some of these things will eventually reach a level of legal scrutiny which might be out of all of our hands. From a technical point of view, we need to develop this information first.

Another major issue is verification of results by a second analyst. That is required in laboratories. How are we going to be dealing with that on-site? Another major issue is adulterant detection. Is this going to be enhanced? Are we going to have similar capability and what can we do along those lines?

The next areas do follow the old checklist. What will be the nature of a laboratory certification program, how and whether, to what degree collection sites will be certified and reviewed? Is that necessary, and if so, how will that be done? The PT issue is another major issue for these kinds of specimens, laboratory or on-site inspection program, the ability to do blind samples.

We have a lot of these issues which may appear obvious when we have dealt with a laboratory, but the Board is going to have to deal with how these can be effectively done or matched in an on-site program.

Finally, and there are probably not a lot of issues here with regard to reporting, certifying scientist type reports, is that going to be done on-site, is there going to be some other way to do that, is that comparable to laboratory, and the items that are in the reports are relatively easy, the specific drugs, can it be done timely and can there be a format.

On the last page, we have again some information about the role of the medical review officer in interpreting results. Here there may be some significant things, the nature of these tests, the conduct of the tests, the conditions under which the tests are done, which are now outside the laboratory, may put a larger burden on MRO's to understand this type of technology and deal with it, and their training and involvement in this process is another issue.

The last miscellaneous issues are again just coincidental to all of these things.

We have re-created, without trying to answer the questions at this time, this is again the guidance checklist that we are using. We are going to take this information back to the Board, and run the information that we have through the checklist. Therefore, we urge people to provide the maximum amount of information to allow the understanding of these things.

One of the things we did when we set up this meeting was to notify all of the laboratories and known manufacturers, suppliers, and distributors of products. We asked them to make presentations this afternoon. To facilitate this process, we sent them a letter with a series of questions. These are the questions that were asked of the manufacturers, and they somewhat follow this new checklist. This is what we are seeking from those parties, laboratories, manufacturers, others who have experience in these programs. These are the things that we have to have as information in a comparative sense to look at this on-site testing thing.

I would like to point out that I'm going to encourage the manufacturers to provide information at any level. We realize that today, in each product, everybody may not be able to answer every question, but certainly that information should be forthcoming. If it is available today, some of the members of the Board that are here and our staff are going to try to make notes during the presentations to see what information we have on these questions and try to link these together for the future. For those people that will present later, as this information becomes available, if you don't have it today, we would like to have it, even if it's in a concise one line form or in a more detailed summary.

We also recognize that there is a lot of proprietary concerns, and a number of people may not want to say something or speak because they are afraid they are going to give away secrets, but in the end, the Board is going to make the regulations. The degree to which the nature of the various products and concerns are covered in those regulations will be facilitated by how much the Board gets to know about what the ideas are for these kinds of products. It is in everybody's long term interest to make sure that the regulations best reflect this process.

The questions are as follows:

What are the testing principles? Right now, it appears it is still immunoassay. Is it the same kind of immunoassay, do they have the same characteristics, and are there other principles.

We are not necessarily at this time limited to immunoassay. The original guidelines in 1987 limited us to immunoassay. We have the opportunity to change that type of thing in the next set of guidelines. If there are other ideas out there, we want to make sure the guidelines incorporate other capabilities.

We need to know the performance characteristics of these devices. We have asked laboratories over the years in the regulative process to give us their performance characteristics. This information is going to be critical to this process.

What these standards are set out are yet to be determined, but will certainly need to be compared to the laboratory type of standards. If there are scientific and technical issues which need to be involved or encumbered in this, things that maybe we have not thought of that would make this kind of process work, we have to consider them up front. We also have to consider how that impacts on the laboratory program.

Specimen treatment and handling. Since we are talking about both oral fluid and urine, we need to have those kinds of recommendations.

We are going to try to decide whether these products will have a FDA clearance requirement, and therefore, we do want input from the manufacturers on the feeling about the FDA process, whether that is useful or whether it should be required in the things that we do in the future.

The clinical studies do demonstrate the product efficacy. We are hoping to see how this compares -- you will hear later this morning from Dr. Willette and others about studies that have been done focusing on product comparisons, but exactly how did you do that or how do you see the clinical efficacy of these products.

A big question is how do you recommend dealing with adulterants. Are these products going to be used exclusively, do you recommend adulterant testing, will adulterant testing be inherent in any of the products.

One of the biggest questions that is on this list is on the next slide. How should the device be evaluated prior to use. This is probably a very fundamental question. I know this is an area that has been of a lot of concern in point of care type clinical testing and differences from CLIA recommendations and things like that, and there have been some exceptions to these products made, but on a general basis, batch to batch variations and daily prior use is a critical component of all types of forensic testing.

How do you recommend doing this? This is something where we need to know what the users' experience or recommendations are. Again, if you go back to the things we mentioned earlier in re-reading the guidance to ensure a forensically satisfactory process overall.

We would like to know what the unique features of various products are. We know a little bit. I think hopefully we will hear a lot about that this afternoon. Are they merely immunoassays done again, reborn on a strip of paper or are there other unique features about these products.

Quality control recommendations, external PT recommendations, training recommendations, of particular importance. We have limited parallels for training of this on-site thing that come from the DOT alcohol experience. I think there have been some concerns about that. The concerns have been mitigated by the fact that the DOT alcohol is a very limited program. It didn't go broadly to pre-employment testing. Therefore, it has not been utilized to the degree, and the incidence is relatively low. How is that going to affect widespread urine testing where the arena is used to a laboratory type of scrutiny.

Finally, there are a couple of other questions. The last one that was sent out was how do you recommend the use in the workplace setting. That is a system. We really like to hear about systems. Are we merely talking about products or are we talking about systems.

Somewhere in the course of the day, we want to be able to differentiate products from systems. Right now, we do have products in the laboratory, but the laboratory is a system. We are going to have on-site devices which are products, but they are going to have to be used in some general context, which is a system.

The final things were not on the original send out, but just to help us try to figure out the comparison of these products, we would like to know if you are a manufacturer of the product, making it yourself and are responsible for the control of the product development, or are you repackaging another product. It would help in looking at how many products are really out there or how many different systems are really out there and whether they are truly different or not, and if we know in general terms the source of the materials.

Those were the questions sent out and these are the things that we need the information on that we hope to seek starting today, after today, and to continue this as part of a process.

Parenthetically, I would just say for those of you who are more involved with some of the other areas, we have already done a lot with the hair testing with multiple working groups. The oral fluid group has met several times already. The sweat patch group has met several times. They are close to the point where you reach a limiting point where input is available. Then at some point, which is going to happen mostly next year, there will be more and more of the Board's assimilating these things. We are trying to bring the on-site issues to that same point so that some time early next year, we will have input from each of the areas and that the Board will more systematically during the course of the next year be able to look through these things, as we have said before, both independently and across the grid, looking at the products in relation to one another.

A few ground rules, particularly for this afternoon. The morning presentations are structured. We have asked a number of people to come because they have some experience. Following those, as time permits, we will take some questions from the floor and try to deal with them this morning. This afternoon, we have a number of presentations. We set up a number of slots and we have just about the same number or very close to the same number of people responding. There will be 15 minutes for each of the presenters. Because there were some concerns, we did develop a series of ground rules, rather than have someone stand up with their product and be attacked by all their competitors. We are not going to have that.

We did a couple of things. One is that we are asking this afternoon, because we do want input, we want people to be stimulated when someone is talking, to think about not only what are they saying about their product, but what does that mean. If you have a question of a general nature that might need to be posed, where getting the answer would help, we are going to ask this afternoon that you write them down either on some cards that will be given out or some notepads, and as we go, we will collect them and try to collate them into some common areas. There should be some time at the end. Unless there are a couple of new presenters, we will have the last 30 to 40 minutes of time in the afternoon to deal with those questions, and we will group them together and have a group discussion. If there are specific questions about an individual product, you can write those down. We may or may not be able to deal with them, but certainly we would get them back to the individuals and would pose them to the individuals that give presentations so that they can try to answer those questions in the afternoon. Unless the presentations are short, we have allocated 15 minutes to each presenter, and we don't want to try to delay that, we want to give them their whole 15 minutes. To facilitate the process of how this is to be done, by lot, by a very complicated process involving eight pieces of paper and a cup. [Laughter.]

We developed an order of presentations for this afternoon and tomorrow. I hope all the people are here. Since NOTA was the prior coordinator for some of these activities, we have asked them to be the first speaker this afternoon and Dave Evans is going to do that. Following that, there is a list that was determined by lot in the presence of several witnesses. We picked this randomly, rather than try to do it systematically. There are some people concerned as to who is going first and who is going last and that sort of thing. First will be NOTA. The second is Dade Behring. Third, National MRO. Fourth, PharmChem. Fifth, Roche. Sixth, Princeton. Seventh, Chimera. Eighth, MEDTOX. There was space for the last 30 minutes, which we had allocated, which is currently unallocated.

Is there anyone else in the audience of any of the groups who has not told us before but would like to take time? We would like to hear whatever you have about the product. We know that most people don't have the answer to all the questions that are on there. We do want to hear what products are out there and what the general thoughts are. If anybody who has not signed up to be on the "listed program" does want to make a comment, please let us know, and we will let you have a few minutes.

All of the products for today deal with urine. Tomorrow, there are three products that deal with oral fluids. We chose to separate the urine from oral fluid and deal with the oral fluid tomorrow. There are three. The first is STC. The second is COZART and a third is AVATAR. After that, we have another open period of about 30 or 45 minutes. Depending upon the number of questions, you have time to discuss the oral fluid products or any further discussions or we will end a few minutes early tomorrow.

This is the sequence of events that we are going to deal with today. I think we are ready to move forward with a series of presentations.

Let me just take one minute, does anybody have any quick questions about today's processes? We want to make sure everybody understands where they are and what's going on today.

MR. FORTUNA (Chemical Detection Services): I'm curious if you would entertain other questions other than the ones you have raised this morning.

DR. CAPLAN: Such as? In what area?

MR. FORTUNA: The first one is when is it appropriate to use a particular kit? Would you use it for pre-employment, use it for random?

DR. CAPLAN: I'd say the answer to that is yes, as time permits later on this afternoon. The discussion will not be limited only to the questions we posed. Those are mostly for the devices. As time permits, we want to use the time to discuss anything that will be associated to help us better understand or write the rules. The answer is yes, but it will probably be late this afternoon.

DR. CAPLAN: Dr. Robert Willette is going to give us the first discussion on performance of non-instrumented drug test devices.

STUDIES ON PERFORMANCE OF NON-INSTRUMENTED DRUG TEST DEVICES (NIDT)

DR. WILLETTE (Duo Research): I see many familiar faces in the room that have been involved in on-site testing for the past 30 years. On-site testing is certainly not new. What is new is the evolution of the Government regulated drug testing and question whether on-site testing is suitable to be used in those various forums that are regulated by the Federal regulations. One of the things that obviously is critical to making decisions about in particular the quality control criteria, performance testing, and some of the factors that are in place for laboratory based testing, is based on studies that are done on the devices.

The first thing we need to do is to look at the literature. There are several published studies evaluating drug testing devices, and we are focusing here on the non-instrumented ones. There are studies that have been presented at technical meetings. Certainly, the manufacturers have data and studies they have conducted or commissioned. This information is generally available. Some of that usually is always incorporated in the FDA cleared package inserts that are available with the test devices. There are a number of documents available from the FDA that provide guidance as to the nature and kinds of studies that are recommended, some of which obviously are required. The nature of those studies may vary from device to device, depending on how it is similar to existing products, how it may differ, whether there are unique characteristics that may require studies that may differ from some of the other products. A number of these are available from the FDA. There is a number on their Web site if you want to see the variety of studies they recommend.

The kinds of studies that we see are done in a variety of ways, and they are all complimentary. They may serve different purposes. They may answer different questions. Certainly, a good start is to take well characterized quality control samples over a variety of concentrations to see how the device performs, specimens that are spiked below the cutoff, above the cutoff, and around the cutoff. Clinical studies usually involve randomly selected or every specimen that comes in the door to run these on devices to see how they perform in day to day working situations, whether this is done in the laboratory or in an on-site situation. That data helps to characterize a number of features of the device.

Other studies may focus on specifically selected specimens, and these could be specimens that have been either previously tested and known to fall either completely negative or very high positives, or it could be spread out over the range of concentrations or concentrated around the cutoff. Again, each of these study designs can serve different purposes and provide different kinds of information. Then the question is if you are evaluating devices, what do you compare it against. What is the standard. There are some studies that will compare the on-site screening devices to laboratory screening results. That gets a little more complicated because we know the immunoassays, all commercially available immunoassays used in the laboratories, do differ in their cross reactivities. One has to be careful about knowing what they are comparing the on-site devices to. Perhaps the ultimate answer in evaluating them is to evaluate them against the standard confirmation method, and under the current regulations, that is to use GC/MS. One can do a variety of these kinds of sample selections to evaluate the device, compare the performance of the device then against the confirmation results.

I want to just illustrate the two studies that we have done for two different Federal agencies. The design of this study was to look at samples that were concentrated around the cutoff, and that was for a variety of reasons. One was to see how well manufacturers of different devices can produce a device that has a reasonable expectation of performance around the stated cutoff. It also was to give us some information about what kind of performance characteristics could be required in a regulated program. Do these devices have to have precision that is plus or minus 25% around the cutoff, plus or minus 50%, what really is the expectation based on what is available.

The first study we did was for the Administrative Office of the U.S. Courts. This is because the U.S. Courts and Federal Pre-Trial and Probation sites do a lot of on-site testing.

I don't know if everybody in this room has seen all these devices, but from this first study, these are what they look like. These are multiple test devices. They come in various forms. Single test devices can be cassette type or what we call dip stick. In the first study we had one that was in the cup format.

We did a second study approximately two years later to see what new products came on the market, what improvements perhaps were made, and whether there was any unique or different products available. This was done for the Division of Workplace Programs at SAMHSA. Again, these are the devices that were included.

We also ran the AdultaCheck, evaluating only the creatinine component. Only the creatinine because these specimens all came from submissions from Federal Probation Offices, who do primarily observed collections and the incidence of adulteration is negligible. In the 450 plus specimens that we included in the study, we saw no other adulterants. We were evaluating the creatinine part of that assay.

These were the devices in the second study, the number of cups proliferated. There were four available for the study. Multiple devices take on the form of multiple dip sticks as well as the cassette types, and single test devices, dip sticks and cassettes. The one on the far left is the AdultaCheck. It is currently available and has four pads to do glutaraldehyde, nitrite, pH, and creatinine.

We compared them with an on-site instrumented device, the ETS running EMIT d.a.u. reagents was used, because that was the primary instrument that was used in the Federal Pre-Trial and Probation locations that were being compared to the non-instrumented on-site devices.

The protocol was to test the 15 non-instrumented devices and the instrumented ones side by side, using 90 specimens that were selected from submissions from Federal Probation Offices. We specifically selected specimens for both studies based on the initial laboratory screening results. The contract laboratory for the Administrative Office includes positive and negative controls at plus or minus 25% of the cutoff. We used the immunoassay results from that initial screen to select those specimens that fell in that range, trying to get about 60 specimens, 30 below the cutoff and 30 above.

In the first study, the laboratory is using EMIT. In the second study, it's a DRI enzyme immunoassay. That is one difference between the two studies. The samples were tested on all devices on the same day, approximately 20-25 samples could be run a day on all 16 devices. Then the EMIT d.a.u. reagents were used as an instrumented on-site referenced method. Over the course of both studies, we hired three laboratory technicians. In part of the study, we had two. They were trained to follow the manufacturer's protocols. The samples were coded so they didn't know whether they were negative, high positive, or samples that were close to the cutoff. They did know which drug was being tested because when you are mixing single test devices and multiple test devices, we weren't running every single test on a given day to see what drug was present, so it would be a cocaine day or a marijuana day or an opiate day. Everything was coded. They did not have any expectation of whether it was going to be positive or negative.

We did a single interpretation per result. This was primarily to model what was being done in the field in Federal Pre-Trial and Probation locations. They didn't have double readers.

We also had to deal with what we called equivocal results. Many of you, certainly the manufacturers and distributors of these devices, know all too well, but some of you may not, that what you are looking for as an end point is the formation of a line in almost all cases to represent a negative, and the lack of appearance of the line to be a positive. Sometimes it is difficult to distinguish whether a line is actually present. Those were graded as borderlines. They had to be either borderline negative or borderline positive. We did a lot of data analysis looking at that.

In some cases, if you graded a borderline positive as positive, it helped the performance of the device and with other devices, it actually degraded the performance score, but when you looked at it overall, it wasn't a really significant difference in including bar line positives as positives. The data I'm going to show you actually represents scoring all the borderline results as negative, and that is the conservative approach. Most package inserts say if you see or you think you see the presence of any kind of a line, call it a negative. That is erring in the favor of the individual. The data was all recorded on a daily basis and then analyzed later on.

I'm just going to show you a couple of drugs to give you a feel for what we were looking for and the kind of variability that one sees between the devices. This is from the first study done for the U.S. Courts. These are the amphetamine results, green bars are false negatives, red bars are false positives. Just trying to illustrate the variability from device to device, running the same specimen on all these devices on the same day.

In the second study, the coding on the bottom, you will see that some of the capital letters are followed by a small "a" or a small "m." That is because some of the devices are amphetamine specific and some are methamphetamine specific. I think there is only three devices up there that had both, the amphetamine and methamphetamine test strips included in the device.

We have done some subsequent analysis. I haven't added it to this slide. What we are showing here is if you evaluate these specimens using the amphetamine strip or you evaluate using the methamphetamine strip. One example here is the amphetamine and here is the methamphetamine. See the difference between them. In a subsequent analysis, what we said is if you have both strips available on a device, if either one shows a positive, then you send it to confirmation. When you combine both strips, the number of false negatives drops considerably.

In this particular sample mix, using the HHS cutoff criteria, there was improvement if you could judge both the amphetamine and the methamphetamine strip.

Under the U.S. Courts' confirmation criteria, which has lower cutoffs and doesn't have the 200 nanogram rule, the performance pattern looks fairly different because this population is predominately methamphetamine users. Under the Courts' criteria, there was 29 methamphetamine positives. Under the HHS criteria, only 10 of those met the HHS criteria. It's a different mix of specimens. In both cases, judging both of them, it dramatically lowered the false negative rate. The ETS is the first one on the left. This one on the left is the perfect device which none of you have come up with yet. The one on the far right, this is the AdultaCheck, creatinine, on this set of specimens for amphetamine.

The first study, the cannabinoids, again, considerable variability between false negatives and false positives, running the same specimens side by side. In the second study, again, we still see variability between the devices.

Cocaine and opiates. There wasn't as much difference between devices. There is a little bigger difference in the opiates when you go from HHS rules to the U.S. Courts' cutoffs, because their opiate cutoff is half and they include hydromorphone and hydrocodone, which is somewhat prevalent in the U.S. Courts and Federal Probation.

Neil, how many hydromorphones do you see in a year?

MR. FORTNER (PharmChem): About 30 percent.

DR. WILLETTE: Adding hydromorphone lowers the false positive rate of these devices by a substantial amount, but under HHS criteria, those obviously are called negative. I didn't include the Courts' analysis here, but the false positive rates for opiates go down. Cannabinoids is the same cutoff. There is no difference between the two programs.

If one is using a multiple test result with all five drugs present and whether we had single devices or multiple test devices, we ran all five drugs. If you are looking for overall performance, we averaged the results from all five drugs, and in the first study, there was greater variability, and you will see in the next slide with the devices that were in the second study, whenever you are doing studies, I mean, this is a different set of specimens in the two studies, the composition varied somewhat, we had more amphetamine positives in the second study then in the first.

There are differences in the two studies, but as we will see in this next slide, the overall performance of the devices in the second study improved. There is not as much difference when you average all five drugs together, even though there may be significant differences from drug to drug, when you take all of them into account, they cluster together.

How do you evaluate data? These bar graphs show you false positives and false negatives. It's in the real world, the user that is out there, the collector of the company that is doing the testing, they don't know what the number of true negatives are and they don't know the number of true positives. All they know is they got a positive or negative result, and the question is was that correct or not. The other way of looking at the same data, this is based on taking a GC/MS confirmation of all the specimens and saying, well, there were 30 true positives, but the device only detected 10. That is a 67% false negative rate.

The other way of looking at the data is to go to positive predicted values, and that is saying if I have 50 positive results from a device, how many were confirmed. If 40 were confirmed, that is a positive predictive value of 0.8. That indirectly relates to what the true positive rate was, but the user is more concerned about having the positives confirmed. They do not like false positives, particularly if they are going to take immediate action, removing somebody from a safety sensitive job or whatever. This is another way of looking at the data. Obviously, the perfect device has everything right, but in the first group, when you plot the negative and positive predictive values against each other, you see there is a lot more scatter. It's maybe not as easily distinguished in the bar graphs, but there is a lot more variability between the devices in the study.

In the second study, they have clustered closer together. As pointed out, there are devices, at least one, using the HHS data, that actually showed better performance than the ETS.

The AdultaCheck, overall accuracy, how many of all results were correct, positives and negatives, it was about 92% in being able to distinguish the specimen that had creatinine at 20 or above or below 20.

We think part of the reason the overall performance improved in this second group is one factor was we did get more positive amphetamine specimens in this than in the other one. The fewer specimens you have, when you miss one, it has a more significant impact on its performance. Also, these all came from recognized, documented manufacturers that had a history in the diagnostics field, so it was a more select group of devices, and maybe manufacturing processes improved. There is a number of possible factors that go into this. The analytical performance isn't the only criteria. There are a number of other factors that may affect what appears to be an analytical performance. Ease of use. Certainly, the simpler it is to use, perhaps the fewer problems and fewer incorrect results will evolve. The specificity of the antibodies. As we well know in immunoassays, antibodies are not all the same. When you are using spike specimens to do a comparison in a real world specimen, you are going to see significant differences. Quality of manufacturing. What is the lot to lot variability? We saw a little bit of this in the second study because we had a number of devices that came from the same manufacturers. We could see differences. I think in one case we had four products that came from the same manufacturer. They didn't all perform exactly the same. We know there is some variability either in the manufacturing of lot to lot of the strips or in the subsequent processing to put the strips into the device. Reading results. This is a borderline issue on whether there is a positive or negative result, is there a line present or is it a shadow or what is it. It is certainly a critical factor. Clearly, the greatest distinction between those positives and negatives will affect the outcome. It is one of the reasons we concentrated the specimen mix around the cutoff, to see how well one could distinguish specimens that would confirm positive or not. What are the operator differences? We are going to hear a little bit more about that in a few minutes. Then the population to be tested as you get into the field that Jim Baer is going to talk to us about. They have significantly higher positive rates than in the workplace. The whole issue of false positives and false negatives takes on a different picture. Judging these results. This is an important factor. As I said, most devices -- this is the very simple approach to doing these solid phased immunoassays, have a colored line for negatives and the absence of a line, no line forms, if it's positive. Triage produces the line for a positive result. Triage is the only one that we have seen that has the line for a positive result. The old profile, is that still available, the old immunoassay?

Of course, the instrumented system, EMIT, or whatever reagent is used on an instrument, it gives you a numerical printout which is easier to discern what is positive and negative than in many cases with these lines. As we have tried to illustrate, even using this sort of stress test of concentrating real specimens clustered around the cutoff, there were some devices that were certainly comparable to the on-site instrument. We saw wide variations between negatives and false positive rates and borderline results. As I said, the borderline results varied widely across devices. Some had a high percentage of borderlines and some had a very low percentage.

Looking at the overall accuracy, there wasn't major significant differences. Certainly, some devices were sufficiently accurate to be used in testing scenarios with adequate precautions. As pointed out, one of the purposes of this working group is to learn from everybody that we can, what are those scenarios where these are appropriate, and what are the adequate precautions, going down the factor grid or going through these questions is the kind of information we need to be able to craft a program that is still going to protect the individual that is being tested.

I have to acknowledge the sponsors of the study: PharmChem Laboratories, who gave us free space outside the main laboratory, so we weren't interfering with their operation and vice versa. Dr. Leo Kadehjian, who was the man on the scene every day supervising the technicians and then the technicians that we hired to do most of the work.

DR. SALAMONE (Microgenics): In the selection of these samples, you used GC/MS to identify them as positive and then the ETS as just a reference screen; was that it?

DR. WILLETTE: Yes.

DR. SALAMONE: The positives that you chose for GC/MS were screened with the EMIT system?

DR. WILLETTE: The specimen selection was based on the daily screening runs at the laboratory. In the first study, the laboratory was using EMIT, but in the second study, they were using the DRI enzyme immunoassay. We took the screening results and selected specimens that fell in that plus or minus 25% range and those that exceeded it and those that showed no response, so that we had approximately 15% that were negative, 15% that were high, and 60 percent in the middle, based on the initial screen. They were tested on the devices and concurrently in many cases, they were being GC/MS, but whether they were negative or positive in the initial screen, they went to GC/MS. We had GC/MS data in all specimens in the second study and almost all in the first study. There were a few that couldn't get GC/MS. They weren't included in the data analysis.

MR. WILLIAMS (Doctors Laboratory): Isn't it correct that a true false positive is the result rendered by a totally negative sample, so therefore, very few of your false positives were true false positives?

DR. WILLETTE: Yes. We have a lot of data analysis on that but time didn't permit including all of it. There is another way of looking at the data, how many of these specimens had drug present but may have been below the cutoff. We have actually done what is called ROC plots. They are sort of the opposite of what you normally do to see what is the optimal confirmation cutoff for a given device. Obviously, the lower your cutoff is on a confirmation, the better performance of the device and false positives drop dramatically. There were obviously some specimens that had no drug present, and that was most characteristic of the amphetamines. There is still a fair bit of cross reactivity in the amphetamine devices with over the counter drugs, and the other problem that I mentioned with opiates is oxymorphone. We had a number of specimens of oxymorphone. Most devices picked it up, but under HHS rules, those were negatives. That is all in the papers in a lot more detail. The second study, the names will be presented next week in Puerto Rico.

TWO READERS OF NIDT RESULTS - CONCORDANCE AND ACCURACY

DR. KADEHJIAN: I provide oversight of all of the on-site testing programs for the U.S. Federal Courts. I am also on the faculty of National Judicial College where I teach a class on substance abuse and neurobiology of addiction and other drug testing issues. Because of my experience with Bob in actually running these last two studies where we looked at 30 devices with several thousand specimens, given that experience with these non-instrumented devices, I was asked to do a commissioned study by a vendor of several non-instrumented devices, and what distinguished this study that was commissioned from the previous two was that we actually had duplicate reads of each specimen's results. In the previous two studies, we had multiple operators but each one ran their tests independently, so we did not actually have duplicate reads of the same device. Although Bob did not show that data, we found that there were very little operator variabilities, although some operators reported more borderline or inconclusive results than other operators. We noticed that their overall performance was virtually the same. Again, that did not involve duplicate reads.

In this commissioned study, we actually had each device read by two readers, Donna was very interested in the data on concordance, that is how often did the readers agree, how often did they disagree, and how did performance change when you required two operators to have the same result. One proposal might be to mandate that in order for somebody to be called positive with an on-site result, that two readers have to look at the device. This sounds pragmatically unnecessarily demanding, but nonetheless, it would be wonderful to have data that would address that issue right up front. That is the data that I am going to present to you.

This data comes out of this commissioned study which will be presented at the SOFT meeting next week. I'm also not here to promote any particular devices. The slides I will show you do have device names. It's not important for our purposes today. I just didn't bother making slides that did not have device names.

With that caveat, I will show you the first slide. I just want to point out part of our specimen selection. Many people in the laboratory community are very concerned that devices will not perform well, non-instrumented devices will not perform well, because of the subjective nature of the read. People being unable to decide if a specimen is positive or negative. We wanted to make sure that our studies truly addressed that issue. Although many manufacturers may say we tested 100 negatives and we tested 100 positives and we got them all right, they don't necessarily tell you what those specimen concentrations are. The studies that we have done with Duo Research include both negatives and positives, so we tried to weigh our studies with a specimen population that is truly weighted around the cutoff. We well appreciate this is not the true clinical scenario. We don't have a lot of data on what the true clinical scenario is. That is what proportion of specimens coming from various sectors, whether it is an E.R. tox setting, a workplace setting, a corrections setting, what proportion of those specimens truly are around the cutoff and will be equivocal and perhaps difficult to read. The only data that I found in the published literature on immunoreactivity, not GC/MS confirmability, but immunoreactivity, which is what we are concerned about with these non-instrumented devices, was a study that was done by Roland Valdez, and he noted that in the E.R. setting, less than 10% of all specimens were within three standard deviations of the cutoff; less than 10% of all specimens, and this was several thousand specimens. He did notice some blip at higher concentrations.

We well appreciate that our studies are unfairly weighted against the devices by being weighted around the cutoff, but it is to address the laboratory concern about how well do these devices do. Another thing that Bob did not show you was we actually looked at the devices' performance with concentration. That is how well did the devices perform with specimens that were clean, clean negative. How well did they perform with minus 25%, plus 25% and higher. As you would expect, you would see the devices with negatives generally had higher accuracies, as you move closer to the cutoff, the accuracies drop, and as you get to higher concentration, the accuracies go up again. This is also true for the instrument, the bench top analyzer.

With that caveat, I want to show you data from the current study. Just so you appreciate how challenging this specimen set is, I have provided some concentration. I just want you to take a note at the concentration of the specimens that we are looking at. These are all very closely clustered around the cutoff. We have a certain number of negatives, that is below the GC/MS criteria, but we have a smaller percentage that are truly clean, that is nothing by immunoassay, nothing by GC/MS. Notice that the highest concentrations that we have are only on the order of 1,000 to 2,000 nanograms per mL, truly challenging specimen sets.

This shows the accuracy of the bench top analyzer, about 80% overall accuracy with this challenging specimen set. The accuracy from the scientist who read the devices and the accuracy from the non-scientist. I will assure you that this non-scientist was a non-scientist. I hired a scientist, basically a Bachelor's level person, who was going onto graduate school to get a Ph.D. in chemistry. This person understood all the rules, knew how to do laboratory testing. The non-scientist had no laboratory training or scientific training whatsoever in their background, and the first day they showed up for the study, I thought the study was going to be ruined, this person is not going to be able to follow these instructions. Yet, at the end of the study, this person had accuracy that was comparable to that of the scientist. In my view, no significant difference.

I also want to point out that the number of borderline results, that is how often does the reader report a result as borderline, not very different for the two operators. In fact, the scientist was more prepared to call a result borderline, that is unclear. Remember, we want to capture the information for when these devices are giving absolutely clear results, clearly negative, clearly positive, or I'm not sure. Again, we did review results with both conservative reading for package insert, all borderlines negative, or aggressive reading, all borderlines positive. I only show you this to just give you some idea of the differences we observed between operators. Even though there were differences in the rate of reporting of borderline, again, with the scientist being more prepared to discern those differences, you saw in the previous slide that their accuracies were virtually the same nonetheless.

DR. WALSH (Walsh Group): Did the scientist and the non-scientist look at the specimen results at the same time?

DR. KADEHJIAN: Scientist one ran the test. As soon as the test was ready for reading, they wrote the result down. The other reader came over and read that same result. There was a three second delay, ten second delay in read, but read that result independently. This shows the overall accuracy for negatives in terms of negative predictive value, the overall accuracy for borderline, and they are called borderline negative. That is per package insert instructions, the conservative read, when it's borderline, you are supposed to report it as negative. When that happens, what is the accuracy of borderline results, and finally, the adequacy of positives. Even though the laboratories may say this is a subjective read and we are not going to have good accuracy from these devices, you see that even when the result is fuzzy, uncertain, if you will, borderline, when those are called negative, the accuracies are comparable to the accuracy of clear negatives and clear positives.

This is the concordance part. What percentage of results were called negative by both readers for all devices, all drugs? What proportion were called borderline by both readers and what proportion were called positive by both readers, and then these are the discordant results.

Only about 20% of all results with borderline specimens now, remember, you saw what my specimen set looked like, highly challenging around the cutoff specimen, only 20% of all results were discordant between the two readers. This just shows how it varied from device to device. I don't think this is important other than to note that for this device, this small percentage, they agreed with negatives, borderlines, and positives. You can see that the agreements varied from device to device. The percent discordant results was pretty much the same for all devices, about 20% of all results that were discordant seemed to be relatively independent of device.

This shows the accuracy of result for when they both agree on negatives, both readers call a device a result negative, both readers call the device borderline, both call them positive, and how accurate are the reads when they are discordant? That is when one person calls it negative and one calls it positive. You can see that even though there is discordance, generally, the discordant results are equally accurate to whether they both agree that they are negatives or both agree they are positives or both agree that they are borderline, for the most part.

If we ask, is having two readers agree going to improve our accuracy, the answer is only slightly. This shows the negatives from independent reads. If we take all the negatives, whether they agree or not, what's that accuracy, and then this is the accuracy of the negatives when both readers have agreed it's negative. There is some slight improvement. For borderlines, it doesn't matter whether you have a single read or a duplicate read, they have the same accuracy. For positives, positive predictive value increases a little bit, not by a lot. Again, I have only demonstrated the accuracy of discordant results, when they disagree. Having agreement does not improve the accuracy to any significant degree. This just shows the overall accuracy for single versus discordant results for all devices. It's the same for all devices. Virtually no difference whether you have a single read or duplicate read.

This shows the negatives. They all show some improvement in negative predictive value if you have a duplicate read instead of a single read.

This shows the borderlines. Again, very little improvement, very little difference.

Then for positives, some slight improvement in positive predictive value across all devices.

Is there any rationale for requiring duplicate reads? I think this data says the percentage of discordance is small, less than 20% of all our results. Remember, these are borderline specimens, less than 20% were discordant, and the accuracy does not improve significantly by requiring two readers to agree on whether a result is positive or negative or borderline.

MR. OKORODUDU (University of Texas Medical Branch, Galveston, Texas): You did readings. What about duplicate testing? The inference we are making from this is the non-technical person is equivalent to the technical person. You would think there would be a difference in the accuracy if we are giving these test strips and the same samples to the technical versus the non-technical.

DR. KADEHJIAN: In response to that, the first two studies we did, which Bob spoke about, had that very protocol included. That is each specimen was tested by different operators across different devices. Although no one specimen was tested twice by the same device by two different operators. In all three studies, we found that there is really no significant operator to operator variability in performance. Really, they had no training. Generally, the manufacturer would make sure that the operator had training on use of the device. All we did was demonstrate the devices, provide them the package insert and then ran some control specimens so they had familiarity with the protocol. Understand that in these studies, the operators were challenged to try to remember the operations and reading for 15 different devices, and in my last study, five devices, whereas in the real world, an operator would have one device for the most part, and be running that device all the time, so they would develop a lot of familiarity with it. My experience with these three studies over several thousand specimens and 30 devices is that there really is no significant operator to operator variability, whether they are technically trained or not. It's testament to the technology, that the devices work as well as they do, much to my surprise in reality.

PAROLE AND PROBATION EXPERIENCES AND MODEL PROGRAM

MR. BAER: I have been asked to share the experiences we have had in the Central District of California with non-instrumented drug test devices. Our District is one of 94 in the United States, Virgin Islands, and Guam. There are four separate Federal districts in California. Ours is headquartered in Los Angeles, seven counties around that. We go from urban, suburban, rural. We have 5,000 offenders under supervision, and out of those, about 1,100 are subjected on a regular basis to substance abuse monitoring. We have 29 collection testing locations comprised of probation offices and after care vendors who are under contract to us that provide testing and treatment services, both outpatient and residential.

As a way of background information, I first became aware as the substance abuse coordinator for the Central District of the development of these non-instrumented devices in the late 1980s. Several came to my attention that were being marketed. After looking them over, it was my opinion that they took too long, there were too many steps, they were too expensive, or in the case of one, it was less expensive, it required refrigerated reagents, and we didn't have that capability at the time to provide for refrigeration. In approximately the early 1990s, there was a second generation, as I refer to them, of devices that came out. We began to look at those seriously. In March of 1993, the U.S. Probation Office, our national office, the Administrative Office of the U.S. Courts, Federal Corrections and Supervision Division, which it is referred to now, under the guidance of Ron Dyson and Aaron Lucas, gave us guidelines for the use of these devices nationally.

I indicated that the primary utility of these devices would be as a monitoring and counseling tool. Each district was required, before they could use them, to develop guidelines and procedures. We were to have training from the manufacturer. We were to notify our local district court that we intended to use these non-instrumented devices as a presumptive screening tool. We were allowed to purchase them, but we were required to have them using the same cutoffs as the National Laboratory which we had traditionally been sending our specimens to. The positives that we were obtaining on-site were to be screened by the National Laboratory following its normal protocol. That is using a NIDT for screening and then GC/MS for confirmation. We were not permitted to report any of these non- instrumented on-site positives to either the court or the parole commission unless they were confirmed. They were not to be used solely for adverse action. However, if an offender admitted upon the development of the non-instrumented device to the positive and the officer had no plan to initiate any violation, then the officer was permitted to take administrative action, increase counseling, increase testing, in essence, an informal response to the on-site admitted positive. One last requirement was that 1 out of 40 of the negative on-site results had to be sent to the National Laboratory for independent testing.

In September 1994, it was suggested by our National Office that we be aware that products were now being given an FDA 510(k) clearance, and that we could use that as one of the additional items in terms of selection. In September 1994 also, we made our first non- instrumented purchases. As mentioned earlier by Bob Willette, Duo Research conducted its study in 1996 and released it in 1997 to the 94 Districts. In 1998, we were able to start obtaining these devices off the GSA Schedule. In October 1998, reversing the earlier prohibition from March of 1993, we were now permitted, due to the increased credibility of these devices, to send a non-instrumented positive result directly to the lab, bypassing the NIDT screening and asking for a GC/MS confirmation. Within the Central District, we early on developed a written policy and procedures for probation officers and our after care vendors. Such requirements that before using any of the devices, they had to go through a training program either from the manufacturer directly or from one of our officers who had gone through a train the trainer program.

As you can imagine, many of these devices are quite similar, so as we began evaluating various products and purchasing them, there was no sense in bringing the manufacturer back in if the device was a clone of one we were already using.

Certification. We do not have a formal certification program. However, one is pending, thanks to the assistance of Howard Claussen from MEDTOX Lab, we are now in the process of developing a test and a certificate for all of our officers and after care vendor staff.

I will be referring in a couple of minutes through other slides to our selection process for purchasing the non-instrumented devices and also using a bar code labeling system. We are also participating in a quality control system, which I will also discuss in a minute.

Through the years, we have looked at purchasing the devices and developing a system of preferences in purchasing them, so we have developed a listing that we use on a request for price quotes. First, that it have FDA clearance, that the device be in a sealed pouch, clearly labeled with the product name, the drug or drugs being tested, the lot number and expiration date. For a cassette device, that the pouch which the device comes in must have a disposable pipette, as opposed to a separate bag. An early experience was we would be buying these devices and officers or counselors would be reaching in and the next thing you know, the whole bag would be on the floor or spilled all over the place, so it didn't really lend a lot of credence to this idea when the offenders came in that we were providing a nice sanitary, clean approach to doing the testing.

For dip stick devices, it was our decision that we would not have a separate container to obtain the urine or pour off into. We wanted to restrict it to actually dipping it into the bottle to use to send to our National Laboratory. Clearly, it must have a built in control, be an one step process, no reagents, no washes, at least an one year shelf life from the date of receipt, require no refrigeration, have a documented record of accuracy and reliability. As you can understand, these studies are good for one point in time for the lot that was analyzed and despite that understanding that there might be variability in the manufacturing process, we still require it. We want to see some data. Just recently we added that we would like to have a certificate of analysis for each lot for which the manufacturer sends of the product. I am finding it a little difficult to see that there is a consensus among the manufacturers as to how to go about doing this.

We will be looking at ways maybe to give some more specific information in the request for price quotes in order to have a way of pushing back towards the manufacturer a greater sense of responsibility of verifying for that particular lot that the quality is there.

DR. SALAMONE: If these tests are FDA approved, there is certain criteria that the manufacturers have to meet before they can ship it out. I don't really see where a certificate of analysis would give you any more comfort in terms of the product, because the FDA criteria is pretty stringent for these products.

MR. BAER: I'm sure the requirement is there but I'm not equally sure that they are doing it on each batch or each lot that I receive, and I would like to have the added assurance. Plus, in looking at 15 different devices myself, I've seen a real variability of quality, not necessarily within one manufacturer's product, but over several that had FDA approval. All other things being equal, of course, cost is an important consideration for us. This is the last criterion that I look at. The manufacturer must present in his proposal to me that he has satisfied each of the prior criterion listed, and then I look at cost. We have developed a unique approach to trying to track the devices that we have. It particularly came about when we decided to start distributing thousands of these devices to after care vendors, these companies that we deal with that provide the treatment and testing for us. One of the concerns was that some of them were doing them for their workplace testing clients and were buying the same product, and we were concerned handing over to them inventory which could easily be siphoned off or used for other purposes.

One of the things we did was we obtained bar codes, consecutively numbered, which were then placed on each device prior to the distribution to our probation offices or to our vendors. That bar code then goes back and is tied to the product, the lot number, the expiration, anything that we want to know about it. The tester then takes one of those bar codes off the product, puts it on the chain of custody form and the other one goes on a consecutive log for that site. It gives us some added assurance also when we have seen how some of the testers in using our singles go to them quite rapidly and unless they record what device they are using on a particular offender, right then it gets confusing because we don't try to allow them to do a drug of the day, like run all cocaines today on everyone and all opiates the next day. We want them to customize the testing, and when they are using singles especially, it might be cocaine for this offender, methamphetamine for the next. It gets pretty easy if you are not recording which device you are using to forget. By putting that bar code on, even if they do forget what they tested for, we can always go back and tell them exactly what the product was and what it was being tested for, whether it was a single or a multi-paneled.

It is also helpful for us for inventory purposes because we know what we shipped them, what they ought to have on hand and what they used. Also, if there are any problems of product that are out there and I'm getting complaints back, I can ask for the number that was on that device and easily look it up, and if I'm getting them in from several, I can either contact the manufacturer right away or pull a recall if necessary, which I haven't had to do yet.

Beginning January 1996, I contacted Dr. Willette and indicated to him that our volume had increased appreciably using the non-instrumented devices and that I would like to have the opportunity to have some PT, some external way of also verifying in addition to the 1 out of 40 negatives that I was sending into our National Laboratory, that the devices that we were using had some quality to them. Since that time, we have been receiving 12 blind quality control samples on a quarterly basis. As you can see, in November 1998, we had our first false negative of THC, both by a single and a four panel, and then in February 1999, we had our second, another THC false negative. The single was negative. However, our four panel that we were using was positive for the THC. We have had no false positives on those samples since we started the program approximately three and a half years ago. We have used the samples typically after we have received a new lot purchased from a manufacturer. We have used it also on devices that we have chosen to evaluate.

These are purchases that we have made through the years. Beginning with the first fiscal year, the purchases were actually made in September 1994 and not really utilized until the next year. You can see a progression from 4,400 the first year to this last year, which just ended in September, of almost 52,000 devices being purchased. Those were pretty much spread throughout the year. I try to buy on a quarterly basis, not exactly just in time, but I don't like to stockpile a large number of product, and I have to bar code label them, which adds an additional expense, and I have to phase that in with the personnel that can do it for me.

Also the 40,000 in fiscal 1997, a large number of those were purchased in September, so those probably equalized out with the following year in terms of being somewhere around 36,000 for each of the two years. The composition of my purchases probably has gone fairly close to this, and it's an estimate. Anywhere from 35 to 40% have been multi-panels. We began with the SAMHSA 5, buying five panels. However, an analysis of about 70,000 urines with our National Lab over a number of years, we found that PCP was 0.2% and it didn't make any fiscal sense to continue to test on a regular basis, so we dropped to a four panel. The primary drugs of abuse in the Central District are cocaine, methamphetamine and marijuana. We buy predominately for those singles. Very little opiate and PCP. We pretty much know the zip codes where we can anticipate getting PCP positives in our District, so we buy singles and distribute those to those locations.

Looking at about those 70,000 urines, we found that over a five or six year period, the number of samples that had multiple positives doing the basic five panel screening was 0.7%. Not having this fear of multiple positives, we were able with greater confidence to look at using singles in testing individuals. Over this approximate four year period, you can see the dramatic change that has occurred in our District from approximately 71% being sent to the laboratory to now approximately 27%, and the non-instrumented making up at least 73% of our initial screens.

The consequence is quite stark. Our expenses at our laboratory have dropped quite significantly, from over $400,000 in fiscal 1996 to this year, although my September invoice has not arrived, I'm estimating we will be spending about $260,000. The non-instrumented purchases have increased but not all that significantly. The decrease for the lab has been 38% over that stretch of time, and the increase in our non-instrumented purchases has been 43%. The overall drop in our expense for testing is 24%. At the same time, we have had a 53% increase in the number of offenders that are being monitored.

We collect about five samples, four to five, per month, per individual being monitored.

In my opinion, we see the non-instrumented devices being advantageous because they are easy to use. It doesn't require a lot of training for staff. We have the ability to customize, to test for one or a panel. The quality has been proven to be acceptable for valid screening. For this point of contact testing that we do, for our perspective in the criminal justice system, there is a very valuable advantage to having negative samples being read by the counselor or the officer, because the majority of our offenders are drug free, and the ability to be affirming is a very therapeutic activity, and all too often when a sample is sent to a laboratory and comes back negative, it's forgotten. This ability to say, even though it's always conditional, he may be using mescaline or Ecstasy or something else not even tested by the five panel, we know that we are able to be affirming and to give them a pat on the back, which a lot of them don't get, because so much of our time is spent with the very few that are using.

It is presumptive evidence of use or exposure and it helps us for early detection, questioning of the offender about the positive, and to be able to intervene. To have to wait while someone is a violent predator, a bank robber, and waiting to send that sample off and to come back when we have some immediacy to challenge the person, it is unbelievable, to sit down and watch that test develop, look the individual in the eye and have him respond, what can I say, I'm using again, and you are able to intervene, this is very impressive.

It's not what we are talking about for the workplace, obviously, but for the criminal justice system where we have the obligation to defend the community, to keep individuals who are relapsing from committing new crime, it is an extremely valuable tool, when it is done correctly. In our written policy and procedures, we do not encourage any of our counselors or officers to do any on-site testing if there is any fear that the individual will go ballistic and be confrontational.

Some of the limitations as I see it with the non-instrumented devices, the performance of the devices varies. However, for the devices that we have purchased, I see very little variability.

In my conversations with officers nationally, I am personally concerned that adulteration may go undetected where the devices are used predominately. I don't see much quality control being done in the criminal justice system at all, other than the 1 out of 40 negatives that go in. Our concern with the high positive rate obviously is that it has increased costs, but on the other hand, we can't really afford to have a high false negative rate. We can't afford to have individuals continually using and going undetected.

Data management is an area of real concern. I don't have a lot of accurate information on the use of the on-sites at this point, especially when it comes to the negatives. Out of about 3,000 plus we are collecting a month, probably 97 to 98% of those that are positive go to our national laboratory because we are concerned about keeping an accurate tracking of what's going on with the positives. We are trying to build into our statement of work for our vendors that they develop more sophisticated automated collection systems for the data, for the non-instrumented.

Kind of a minor issue, but nevertheless, it surfaces around the country. Collectors, whether probation officers or counselors in treatment programs, they don't like to test on-site normally. They don't feel it's in their job description. It might be an exposure to infectious disease. They certainly fear some of our offenders, rightly. They mistakenly believe that there might be some reliability issue. The devices we use do not show under the influence, merely the exposure or past use of it.

I think one of the major issues that many of the officers and counselors don't like is when they do the on-site, they might have to deal with the issue right now rather than being able to plan to do it at their own time conveniently in the future when it comes back from the laboratory.

Some of the other considerations that I have regarding data and what's been happening with sending off so many non-instrumented results, both the negatives and positives, what I see happening is that data from the comprehensive labs will be less reflective of drug use prevalence and patterns as more on-site negatives are not submitted, and some on-site positives are not sent for confirmation because they are being handled informally. Data from the users of these devices may be more difficult to obtain because they are not automated. Instrumented systems have a distinct advantage here. With the wide range of quality among all these non-instrumented devices, the validity of data may be adversely affected. With the flexibility to test for one drug or a panel of drugs on-site, it will be necessary to know the testing protocol before readily accepting prevalence estimates or proof of abstinence. Allowing the flexibility to customize individual testing saves money, but destroys the universality of the cherished panel test at the comprehensive lab. This could easily cause havoc for researchers and Government bureaucrats who really love to have standard measurements.

Cost conscious programs will increasingly use or be forced to use the non-instrumented devices to decrease lab costs and increase funds for treatment or to obtain more testing or other alternative monitoring methodologies. In fact, hardly a day goes by that I don't get a call from another district asking me about how we do on-site testing and what product they ought to buy.

Comprehensive labs could lose a significant amount of revenue from customers that begin to use these devices, and we have seen that.

Bob asked me to take a look at our ability to confirm the on-site positives and what kind of disparity we might have off the negatives that go to the lab. We require as part of our policy for officers and counselors to report to me specifically any kind of disparity between a positive that's on-site that goes to the lab and is declared a negative, or from an on-site negative that goes to the lab and then comes back positive. In that, we do not count where someone has used a single cocaine on-site and then it goes to the lab and they do the five panel and it comes up there was marijuana. That doesn't really count in that process. Nevertheless, I have probably had a total since we began doing this of about 30 specimens in which there was any kind of disparity and the predominant number have been marijuana on-site positives that were not confirmed, and these were predominately prior to being able to send a sample that was on-site positive directly to GC/MS. I have not had any calls or any notification of any disparity probably in the last two or three months, which does not mean it is not occurring. Some of our counselors and officers may not be informing me, but I can tell you that it's probably very, very small.

I tried to take a look at the numbers at the laboratory, even recognizing that there are differences year to year in terms of the testing pattern, if there was a significant change in the total number of positives being reported out, to try to get some sense if we are sending in all positives on-site, is there a major change in the lab identifying the positives. Clearly, as we have sent more positives in and fewer of the negatives, our total rate at the lab has increased because the total population of the samples is being skewed toward the positives.

One of the problems also looking at this data is we had back in 1993 to 1995, a seven panel without marijuana. At that time, our positive rate was about 5.3%, with a national positive rate of about 12%. This is a confirmed positive rate. In 1996, we changed over to the five panel, which included the marijuana. Our positive rate then jumped to 6.9%. Back when we were doing the seven panel, we were probably doing 25 to 40% of all samples requesting marijuana, and the positive rate then was about 6 to 7%. During this last period of analysis, our positive rate is about 12%. However, in looking at the number of positives reported by the lab per month, we have gone from about 230 samples being positive now down to about 140 to 170, down to 130 positives, which possibly may be due to our failure to identify true positives using these devices.

I don't think that is particularly likely based on the devices that we have chosen and the evaluation that we are putting them through, both by sending in the 1 out of 40 negatives from the PT samples we are running, and for the independent studies that have been run on these products. It also might be that we are using a poor choice of these devices in testing particular individuals. That is even though we think we know their predominant drug of abuse or the two or three predominant drugs of abuse, maybe we are just picking the wrong singles and we are missing their drug use. I think that is a possibility, but quite honestly, I don't think that would explain the tremendous drop in the number of positives that we are identifying at the laboratory.

It might also be possible that the laboratory isn't performing as well and is failing to identify, but since they are subject to such quality assurance requirements, I don't really see that as being a possibility, unless Neil has some information that he would like to share.

MR. FORTNER: No, we are doing just fine.

MR. BAER: Another possible explanation might be our offender population that's being tested has changed. This is quite possible. There might be fewer individuals using. They have suddenly found the faith. They are responding to our treatment. They are scared. Maybe we have a new crop of judges that has really put the fear into them. It's really hard to know. More of our hard core offenders are in custody and are being kept there for longer periods of time. If you read the research, if you can take those individuals off the street, you have a better chance of lowering crime. It's also possible that our offender population is just using less often.

Another possibility that we have looked at is our collection procedures are less valid now than they were previously, that we are collecting less randomly or they are not being properly observed, that individuals are using the urinator more frequently now and beating us, that they are adding more adulterants, even though predominately we are looking at observed collections. It's interesting to note that when comparing our urine results against the sweat patch for selected individuals, quite often the urines will come back negative and the sweat patch will be positive. If in fact the sweat patch is performing properly, it could be that individuals are in fact beating us on the urines.

In summary, it is my opinion that the non-instrumented devices are effective, they are cost beneficial for obtaining screening information for us regarding drug use, and with their increasingly acceptance by the scientific, criminal justice, treatment and workplace communities, there will be a definite challenge to establish some realistic.

MR. LINEWEBER (UTU): My question is do you use the AdultaCheck strip or have you looked at that possibility?

MR. BAER: We have looked at that product and two others and since probably 98-99% are observed, we see no real utility in doing it. It would just add too much cost. One additional thing that we do in terms of a validity check is that we require all urines to have a specific gravity reading at the time of collection. We use hand held clinical refractometers to test. We have set an arbitrary cutoff of 1.010 concentration. We have had a history of being able to look and have the first sample collected diluted, go to the lab and be negative and a more concentrated sample collected an hour or so later and come back positive.

MR. PATTEN (National Medical Review): Your data showed that the cost of the analysis is going down, the number of tests is going up. Have you seen an increase in terms of the overhead costs, the time it takes for an individual officer or after care provider to do the collection and watch the test and then do the data management?

MR. BAER: Actually, our vendors charge us approximately the same thing to collect the sample and send it into a comprehensive lab as they do to collect it and test it on-site. The majority are negative. They are through with the sample once they have recorded it. If it has to go to the national laboratory, they have to package it up, send it off, get the result back and log it in later. For a lot of them, it's a tradeoff.

DISTRICT OF COLUMBIA PRE-TRIAL, PROBATION AND DRUG COURT EXPERIENCE

MR. CARVER: We do a lot of criminal justice drug testing. Things have changed in the District of Columbia. I'm now head of a new agency which encompasses not only pre-trial services, pre-trial defendants, but also post-conviction probationers and parolees. Like Jim Baer, I'm basically a consumer of many of your products. We have quite a bit different approach, given the fact that here in the District of Columbia, first of all, I don't think we have as good a class of offenders they evidently have in the Central District of California, but we are also much more compact, so we have somewhat different needs, so we can talk a little bit about that.

What I thought I'd do today is just give you a very brief overview of the development of our own drug testing lab. We do have our own lab. We use instrumented testing. I can talk a little bit about how that has evolved over time. I would like to build a little bit on some of the points that Jim Baer made on the rationale for criminal justice drug testing because it's quite a bit different than workplace testing. I am going to run through a little case study because I think it is real important to kind of see in practice how the drug testing relates to criminal justice decision making and what you will be seeing are actual computer screens that the judge sees in dealing with defendants or offenders that come before the court and how the results of drug testing are used, and finally just to cover some of the lessons we have learned in the use of this technology.

We got into drug testing back in 1984 with support from the Department of Justice, primarily to do some research on the connection between drug use and crime, but we really wanted to take that technology and make operational uses out of it. We began with Syva EMIT technology in 1984. The fact that we had set up an on-site lab right in the courthouse proved to be extremely popular with judges. As one might expect, the demands for the services increased quickly. We quickly outgrew the early machines and we upgraded to an Hitachi 705, which did 300 tests an hour. A few years later, we had to upgrade again to an Hitachi 717, 600 tests an hour, and in 1992, we had to develop a laboratory information management system. By then, we were doing hundreds of tests a day and it was just extremely tedious to manually do all the chain of custody documentation and manually enter test results into the computer system, and of course, that left all kinds of possibilities for human error to undermine the integrity of the lab. In 1993, we took what had originally been developed as a laboratory information management system and expanded it to all decision makers within the agency, and then began putting it on the bench in the courtrooms. That became the foundation, kind of the infrastructure, for our drug court.

1997 is an important year. Those of you who are from the District will recall that the District was essentially bankrupt and Congress had to step in and set up a financial control board, and eventually the President stepped in and completely reorganized criminal justice. For this group, I guess the only important thing is that what had formerly been separate D.C. Government agencies, Pre-Trial Services, Probation, which was part of the court, Parole, which was an Executive Branch agency, were all consolidated into a single agency, which is becoming a Federal agency, which I now head. At the present time, we just entered into a new contract with Roche Diagnostics. We are now upgrading to two Hitachi 747's, which provide 2,400 tests per hour, and we also have an on-site to GC/MS. The technical people in here will understand if that has any significance. A reason why we had to add a second lab, we completely were overrun with urine in our original facility.

Just to give you a sense of the volume, in fiscal year 1998, we tested 137,000 specimens. As you will see a little later, this is specimen, and for each specimen, we do a whole panel of drugs, from 27,000 people. We really expect that this number will quadruple or quintuple, and when we get into the rationale for testing and you can see the benefits, I think you will see why. That is why we are building now a second lab. We hope to have that up and running within a month or six weeks. Right now, we have four urine sample collection sites around the city, but we will be expanding that as well. We are essentially moving all of our probation and parole officers into the community. We are getting away from this downtown fortress, supervision fortress probation, as it has been termed. We really want to de-centralize not the lab process, but the collection process, and we are moving toward a sanctions based model of supervision system-wide, trying to establish a level of accountability which our experience has shown can increase treatment outcomes. Key is frequent drug testing, as the measuring stick for all of this.

I know this is a technical group, but I do want to go very quickly over some of the underlining rationale for criminal justice drug testing. You have probably seen this chart before.

Basically this tracks cocaine use over time and it just makes a simple point. While nationally, cocaine use has gone down over time, we have seen a greater and greater percentage of the cocaine being consumed by heavy duty drug users who also tend to be concentrated in the criminal justice system. We know that heavy users consume cocaine at a rate eight times that of light users. There is about 3 million active heavy heroin and cocaine users in the country. About three-fourths of these hard core drug abusers derive a substantial share of their income from criminal activities. I think this has national policy implications as far as where we are going to put our resources.

Studies show that among offenders, high rates of drug abuse are associated with high rates of criminal activity, as one would expect. When you actually look at some of those studies, there can be extremely high rates of criminal activities for the really heavy duty drug users. Conversely, during periods of relative abstinence, criminal activity also declines. This is where the drug testing and the sanctions come in. Legal coercion, in other words, can be effective in enhancing abstinence and improving treatment outcomes. We basically see hard core addicts throughout the court system, not just criminal justice, abuse and neglect, parole, probation caseloads.

This shows drug use among juveniles coming into the court system. Again, when you look at national surveys, it looks like a pretty good picture, but when you look specifically at court systems and juveniles coming into the court system, we are seeing an increasingly higher prevalence of drug use among juveniles. You can see just the progression there from 1990 up through 1996 and in 1997 and 1998, it's even higher, and what is even more disturbing is that you see higher rates of drug use among younger segments of the juvenile population.

It's also a very big factor, as one might expect, also doing drug testing in that segment, although it's not specifically a criminal justice segment.

What we are really saying is that the most dysfunctional substance abusers are already under legal control. You see that "legal control" is in quotes. I think this is not true in the Federal system, but in the state system, state systems around the country, there is really not a lot of the kind of supervision that I will be describing. This group is responsible for a very disproportionate share of the crime and many continue their destructive patterns, despite their involvement in the system. Again, I'm speaking nationally. I'm speaking historically. I'm kind of setting the context of what we are trying to change. It really is a system where we are trying to break this vicious cycle, and criminal justice drug testing, as I see it, there are two main purposes of drug testing. One is the risk assessment, identification of drug users and the risks they pose, and the other is risk management. Again, this is exactly what Jim Baer was talking about in his presentation. As far as risk assessment is concerned, we did some analyses of our own drug test data and basically it showed -- one of the reasons that you want to do multiple drugs is it enhances the predictive power of that information when you are trying to predict re-arrests.

This shows that for defendants who tested positive for two or more drugs, their likelihood of being re-arrested while on pre-trial release, is five times as great as for their drug negative counterparts in the criminal justice system. That is risk assessment. The other part is risk management. Again, drug testing, it's pretty much done I guess in every jurisdiction, but I don't think in my view it has been done with the frequency and the consistency and the broad application that I think would be appropriate. There are large groups that aren't tested at all. Testing is often sporadic and again, I think the system that Jim described is not characteristic of probation testing generally. In many places, probation testing may occur on an once a month visit when the probationer knows he is going to visit the probation officer, and it basically is just not set up in such a way to detect and deal with substance abuse.

With this capability that drug testing brought in 1994, we set up a drug court. It was one of the first ones with assistance from SAMHSA and CSAT as well as the Department of Justice. We did a very extensive evaluation of this drug court demonstration project. Although I won't go into all of the evaluation findings, I will highlight what I think is important in application beyond the drug court, which is what we are trying to accomplish right now.

The idea, like drug courts around the country, and I guess there are 300 or 400 of them now, I think ours is among the first half a dozen to be brought up, but the idea is that you establish this framework of accountability with frequent drug testing, frequent appearances before the judge, and both a system of rewards and incentives and a system of sanctions for people who aren't doing well. The system of sanctions that we developed in connection with our drug court was as follows: There would be very frequent drug testing. Every positive drug test or missed appointment for a drug test, and this is all scheduled, would lead to first level sanction. The defendant would have to spend three days in the courtroom watching drug court sitting in the jury box. If he continues to use drugs after that, it would be a three day jail sentence. Continued drug use after that landed the person in detox, and after that, it was just a seven day jail stint every time the defendant tested positive.

Here's where I'm going to get to the case study. You can almost pick any one of them randomly. The one I'm going to show, I have changed the name, but what I am going to show you are actual reproductions of the computer screen that the judge sees, but it also shows not just a single drug test result, but the pattern of drug use over time. This will take a couple of minutes to walk you through what this screen looks like. Basically, the most recent drug tests are at the top of the screen. The oldest are at the bottom. This is really in kind of a spreadsheet or matrix format where you can see the drugs that are tested for across here, amphetamines, cocaine, methadone, opiates, PCP, marijuana and alcohol. This is produced by our laboratory information management system. I didn't mention it, but it's a completely bar coded system where no one ever has to enter a test result at all. The entire chain of custody is documented and can be brought up on a computer screen. The test results go directly from the analyzer into the computer network and this is basically the end product.

This particular guy, the "L" means this was the date he was locked up, in February, that day, he actually tested negative, but he self reported drug use, so he was put into the testing program. You can see he immediately started testing positive for opiates. It turns out this guy had at least a 20 year addiction career with multiple arrests before he got put into the drug court. He was placed in the drug court here on February 28. One of the eligibility requirements was that you had to test positive at least two times following your lock up test to establish the fact that you really did need the intervention. He is put in the drug court. He now is going to be subject to these sanctions that I described, and you can just kind of watch his progression here. He tested positive, of course, not only for opiates but also for alcohol, and then he was sanctioned and he had to spend three days in the courtroom watching drug court. Again, we are moving forward in time. After that, on March 6, he was all negative. You notice this is color coded. Bad is red. Good is green. You will actually see the dialogue in the court. If a defendant is doing really well and they have these regular sessions with the judge, they will tell the judge, judge, bring up my computer screen, it's going to be all green. It really has changed the dialogue in the courtroom.

Again, a relapse. Addiction is a chronic and relapsing condition. The way it is set up, they are going to be sanctioned immediately. We went to a lot of trouble to make sure that the defendant is back in court the very next day, which of course, is a benefit to on-site testing. We don't have to wait two weeks to get the result back from some lab. As is often the case, the defendant knew he was going to be sanctioned, so he didn't show up. A bench warrant was issued. As also often happens, they know they are going to be looked for, so they will wander in within a day or two. He was spot tested when he wandered in. He was positive. Again, then he did his three days in jail. All this is scheduled testing, by the way. This is not random. This is just a notation that he was unavailable because he was in jail, so that didn't count against him. Once again, he's all negative when he comes out, but then, whoops, he didn't report. Of course, we all know why they don't report because they know what the drug test is going to be. Once again, a bench warrant was issued. He was negative when he came in. There was a sanction hearing scheduled but no sanction was imposed. You may wonder why, this is supposed to be predictable. There is a dialogue going on with the judge and the judge can basically hit a function key and bring up case notes that the case manager recorded. Of course, there is a lot of case management with this, and when you look at the particular notation, he indicated that he had been in the hospital and that fact was in fact verified, so he had a verified reason for not showing up for his drug test, so he wasn't sanctioned. After that, negative. Here, he didn't provide enough for a test. Sanctioned once again when he didn't show up. Here it is a very spotty record where he is consistently not reporting and bench warrants have to be issued and he comes back.

Then we are now, by the way, looking at June and July. We are now looking at five months after this guy first came into the system. All of a sudden, you start to see that turnaround. From then on, this particular defendant basically got it together and again, as is often the case, it was the court pressure that induced him to seek treatment, and here you will find that the case manager put in a note that he was in a treatment program and doing well.

That is just kind of a case study. We have mentioned adulteration. I should say all of our sample collections are observed. When people try to drop things from their fingernails, hopefully the person will catch that. Of course, the biggest thing or way people try to beat the test, and I learned a lot about this from Dr. Kadehjian, is water loading. We have set up the analyzers so that in those cases where you get an urine sample and it's crystal clear, a creatinine test is run. It has been amazing to see how judges themselves have become educated in your field because the computer will automatically put a remark into this computer screen which will say, for example, that the creatinine value was 12. Anything below 20, it will say 170 is normal. I may have these numbers wrong. I'm not a lab guy. Anything below 20 may indicate water loading. You are not only holding the defendant accountable for trying to dilute the urine, but you are also educating the decision maker on a lot of the tricks that people to beat the system.

This basically shows a week by week record of everybody in actually two different tracks of this drug court demonstration project. The guy we showed, you can see over time, he kind of cleaned up, and that's not at all unusual. We actually ran two separate interventions and compared it with a third. I have two of them on this chart. By definition, when the intervention starts, everybody is a drug user. There is 100% drug use here. Over time, you will see that the line goes down and that represents the percentage of defendants who have drugs in their system.

There are two lines here and one of them drops faster than the other one. This thin line at the bottom were people who were subject to sanctions but weren't required to go to treatment, and this thicker line at the top were people that were referred to treatment but they creatinine value was 12. Basically what this shows is the folks subject to the sanctions cleaned up quicker and more significantly than those who were in good treatment but didn't have kind of the pressure of the court.

Basically, just to summarize what we have learned, at each sanctioning level, fewer and fewer people are using drugs and fewer need to be sanctioned. I don't have time to go into it now. I do have some very interesting data on just kind of the progression and how many people need to be sanctioned over time. You see after the second or third sanctioning, a rapid drop off.

Continuing pressure of testing and sanctions encourages many to seek treatment on their own or request a referral. Again, if you are working in an environment as we are in the District of Columbia, which is very treatment resource poor, this in effect serves as kind of a triage mechanism so that the people who have the most serious problems kind of sort themselves into a group that can then be referred to treatment.

Certainly the relationship between the judge and the defendant is very important. What we have learned here is that legal coercion enhances outcomes, certainty and immediacy of the sanctions seem to be much more important than the severity of the sanctions. The evaluators did some interviews with the defendants themselves, and basically what they said was that first of all, they perceived the system to be tough but fair. There was a perception that there was a certainty of the penalties. They knew that if they violated their contract with the judge, that the sanction would be imposed. What was really interesting is they began to internalize a sense of personal responsibility. It wasn't the system doing something to them. It was them not living up to the agreement that they had entered into with the judge with some ceremony, so therefore, they were responsible for the sanction.

Swiftness of the penalties. They knew they had to report immediately after a drug test failure, and that was important. Again, the most important thing is they thought the system itself was fair and they had some control over the outcome. That's pretty rare. That's extremely rare in the Federal system with sentencing guidelines, but in this system, criminal justice generally, it is fairly rare.

That basically is the underlining rationale. What's in store for the future? We are very much interested in taking this model, which works very well for small groups of people, and making that the norm across the system. The system encompasses approximately 30,000 individuals under pre-trial, probation or parole supervision, so it is a big undertaking, but so far, we have had very good support from Congress to fund it, and we do anticipate a quadrupling or quintupling of the drug testing volume in the future.

From a technology standpoint, one of the things that most interests me as an user of the technology is the detection window. This particular example involved an opiate user but when you get into marijuana and if you are dealing with a system that is designed to impose sanctions immediately, and if the literature shows that the detection window for a fat soluble marijuana can be quite lengthy, you can have somebody sanctioned today and three days later, their sanction is over, they test positive again. They say it's from before. We don't know if it's from before. That creates all kinds of operational and administrative problems in an environment in which we are trying to establish this level of certainty and accountability.

I have been very interested not only in advances in instrument based testing, but also in advances in some of the other testing that we are going to hear about this afternoon because it is my understanding that some of these non-instrumented devices, and it may be true for instrument methods as well, can narrow down that detection window so you have a much better handle on determining whether it's residual use, because the guy is a heavy duty marijuana user and is still coming out in the system, or if it's new use and thus a sanction would be necessary.

MS. MASTERSON (LifePoint, Inc): Was there any difference in either the impact of the treatment program or did you track heavy use versus light use?

MR. CARVER: Yes. We looked at re-arrests over time. We looked at drug use over time. Actually, the people in the sanctions program did better in each of those tracks. I have the information here if you would like it. That's the quick summary. I should also say this is a point of contrast between what the Feds are doing with respect to confirmation. I forgot to cover this point. Our internal protocol requires for all positives, we do a second pouring from the original specimen cup and a second test before the results are released by the lab. Any time a defendant or probationer wishes to contest the result itself, now we are able to do it ourselves, but we will send that out for confirmation. I should also say that I showed you that we did 137,000 tests last year. The need to send it out for confirmation maybe arose half a dozen times. I think in part, this is a function of kind of the reliability of the technology itself, and they do kind of accept it as a given. What you will sometimes hear in court is yeah, I'm positive, but I got it from somebody else who was smoking or I got it -- this is my favorite -- when I was packaging the cocaine, it must have got into my skin, but I didn't actually smoke the stuff. That's all protocol.

ON-SITE DRUG TESTING PROGRAM OVERSIGHT

DR. WILLETTE: In the various components in managing on-site testing, whether it's instrumented or non-instrumented, it really has its origins in terms of quality assurance of doing performance testing or inspections with the military back in the early 1980s, and that has been an ongoing program for a long time. We actually began to monitor on-site testing in the mid-1980s in two workplace situations, some of which Leo is going to talk about tomorrow morning. Also, we began in a quality assurance program for the Administrative Office of the U.S. Courts to provide a variety of quality assurance activities, which included inspecting the contract laboratory or laboratories, at one time they had two laboratories, and also a blind quality control service where specimens were submitted through Federal probation offices into the central laboratory to monitor it, and that was coupled with the inspections. Also to monitor the on-site testing, starting back some time in the 1980s, U.S. Pre-Trial Services offices began to do on-site testing in the courthouse or in close proximity to the courthouse to provide testing information for the magistrate or judge before the first appearance.

The Administrative Office, under leadership of Ron Dyson and Aaron Lucas, both of whom are here, we began to extend those PT services and inspections to all of the pre-trial service offices and slowly a growing number of U.S. probation offices that started to do instrumented on-site testing. I'm going to talk primarily about the instrumented program because that has been going on now for somewhere between 12 to 15 years.

The PT samples are manufactured under standards equivalent to FDA GMP's, their compositions checked, they are spiked with drugs for which the on-site operations are testing for, which can include the five so-called HHS or SAMHSA drugs, plus barbiturates, benzodiazepines and alcohol. Most sites now do not test for barbiturates, but we still have a few that do benzodiazepines.

They are stored frozen after manufacturing. They are shipped to the sites frozen. The sites then keep them frozen prior to testing them. After they complete the testing, they are to retain those samples frozen until they are notified by us that they can be disposed of or used for any other quality control purpose that site wants.

Currently, we send 12 PT samples which are coded. They contain a mixture of positives and negatives unidentified to the site. They go to approximately 40 pre-trial and probation sites. I'd say probably 35 or 36 of those are pre-trial and four or five are probation. Currently, the number of kits that are in the current contract, that comes out to about two or three times a year that they get the 12 samples. We used to send them out quarterly, but the number of sites has grown, so it's approximately two to three times a year. The sites test the samples. According to the protocol, they are supposed to test them along with routine specimens that they have collected and are testing. Maybe one a week or two a week or whatever their volume of testing is, and then they record those results. They send them to us along with, in the instrumented cases, the printout of the instrument readings. This gives us a chance to review the instrument performance in the event that there is a discrepancy in the result. We de-code those and then on a monthly basis, we report the results to the Administrative Office and to the site that has submitted results in that preceding month, so the sites get a report, and in many cases, the judges want to see what their performance is. They serve the purpose of helping to support the proficiency of the operator or the operation.

Going back over the last three to five years, the overall score on identifying positives has been 98+ percent truncating. If we get a result that is discrepant and those are any false negatives, we have one or two, maybe three times a year false positives, but for the most part, the discrepant results are false negatives. The site is then notified to pull that specimen or that sample out of their freezer and to re-test it and submit the re-test result to us. Those discrepancies then are investigated during the periodic inspections. Each of the instrumented sites will be inspected on a rotation basis. Currently, with the resources that the Administrative Office provides, frequent than every two years. The program is changing because even some of the large sites like a large probation office that has been doing instrumented testing for a number of years is now switching to non-instrumented systems, so the PT program is in the process of being modified to better accommodate the differences between instrumented testing and non-instrumented testing in the make up of the samples, the size of the samples and so forth. It is also changing in terms of the inspection process, so in this fiscal year, we are evolving an on-site inspection process for the non- instrumented sites.

The inspection process is a key component coupling with the open PT process. I might say open PT. Actually, some of the sites, like one of the large probation offices, actually has to send the PT samples to an off site collection site. They actually prepare them and send them in blind. We have a few locations that actually blind the test operation. That is not quite as common.

The other thing that is emerging are regional test centers, on-site test centers.

I am going to turn it over to Leo who is "the inspector." This is one of the few inspection programs. We had a lapse in funding. The contract didn't get awarded. During the six month lapse, we had people calling and wanting to know when they were going to have their next inspection. When is Dr. Leo coming. He's back in business and back on the road, and he's going to tell you how that process goes.

DR. KADEHJIAN: Over the last years, I've had the fortune to have the responsibility of providing oversight for the Federal courts' on-site drug testing programs. I will take this opportunity to commend Ron and Aaron for the fine work they have done in directing that program, but also to all of the people within the Federal Court system who have displayed to me the highest levels of diligence and professionalism in the way they manage these on-site testing programs.

As a citizen, I commend them for the fine work they do. As scientists and laboratorians, we sometimes shutter at the thought of non-scientists running drug tests. For the most part, these probation and pre-trial service officers don't know the difference between an antibody and their elbow, as an example, but nonetheless, once again, it's a testament to the technology that non-scientists can produce accurate and reliable results as good as any clinical chemistry laboratorian, and Bob showed you the PT results that are 98+% accurate. I think that reflects the diligence by which these people approach their programs. Understand there are no statutory or regulatory criteria for these programs. There is no law that says, if you are going to do court based drug testing, it must be done according to these statutory criteria or to these regulatory criteria. Once again, it's a credit to the management of these programs, to help us have as fine a program as possible. What we have done is tried to impose on these facilities the hallmarks of fine laboratory testing and for the most part, those are blind proficiency testing and external oversight. These are the two key things. Bob is responsible for handling the proficiency test part. I provide all the on-site inspections. I will tell you that the inspections are as much an educational opportunity to these people as they are making sure that they are doing things correctly.

Many times when I'm inspecting the sites, they will actually have me provide a seminar. They will bring in the probation officers, parole officers, and they will have the judges and magistrates. They will actually have me do a presentation.

I'm just putting this chart up to give you an idea of the pragmatic issues of how do we provide oversight of 40 some odd drug testing laboratories around the U.S. As the Division of Workplace Programs thinks about promulgating regulations, how do we provide inspections for such a distributed testing system, and it truly is a challenge. I travel a tremendous amount to do these programs. I try to group them by region so we have several of the Eastern Seaboard, several in the Southeast, et cetera. These just show the dates that I've inspected them. We try to go into each site about every two years. In the same way that CLIA/HCFA has programs that recognize excellent laboratories, once you have proven yourself to have a high level of excellence, you are then allowed to participate in self-inspection. We don't do that. We still go to see each site, but it makes each inspection that much easier when I go in and I know that these sites have performed well in the past, and okay, what are the changes you have made. It makes it easier to go through these. It takes about three hours for me to do an on-site inspection. These are all the pre-trial sites. We have several probation sites around the country in different areas.

The distinction between pre-trial and probation may be the need for urgency of result. Probation seems to feel they have a less urgent need for a result, whereas in pre-trial, many times the results aren't provided at the initial appearance. This is somebody who has been arrested but not even convicted of a crime. The judge or magistrate is going to use the drug test result perhaps to make some assessment of release conditions. In pre-trial, there is perhaps more urgency to have an initial test result available.

I use an inspection checklist. This is it. It is about 80 pages long. It is modeled after the national lab certification program for SAMHSA labs. The goal of this inspection really is to provide them a document that they can use in court when the results are challenged, to say, don't take our word for it, we have been inspected by these criteria. What criteria?

We decided we would model our inspections as close as possible to the national lab certification program, so that they can say here are the gold standards of drug testing and we comport with it. The language that I use in this inspection checklist almost follows verbatim the language that is used in the national lab certification program and the language that is published in the Federal Register, with minor changes. Instead of "employee," some will use "arrestee" or "defendant" or whatever. We make appropriate changes.

Understand, there is really no GC/MS at these sites. It's just the initial testing part. You are welcome to take a look at this. This is proprietary really for Duo Research, but I'll be happy to give you an idea of what this looks like. It is about 80 pages. I want to point out something from the Forward here. We make a point to let them know that this checklist is modeled after the national lab certification program, so when they receive their inspection report, they also get this Forward again. It reminds them that here's where this checklist comes from, so when the zealous defense attorney says, well, here are the SAMHSA criteria, these are the gold standards, what are you doing, and I say, well, here's our inspection, and it comports with these criteria. We also include some criteria from the College of American Pathology because these sites are really not laboratories, formal laboratories, in a true sense, and the operators are not necessarily aware of biohazard issues, so we pulled some biohazard questions out of the College of American Pathology checklist. I am just going to go over briefly the items in the checklist and the kind of criteria that are included, so you get a sense of the scope of what we do here.

There are three major parts. The first part deals with general lab information. Are you doing on-site collections here, are you doing on-site testing. Do you have multiple sites participating in the program. How many specimens do you test for. General information about how the program works. The next two parts deal with collection and testing as separate issues. Collection, how are these specimens brought to the testing facility. They may all be collected at that testing site or they may be collected at remote sites. There will be a central office but there may be remote offices where they are offenders, and rather than having them drive 100 miles to donate a specimen, the officer can collect a specimen there or have a contract facility collect a specimen. That specimen is then shipped into the facility.

We ask about all collection related issues. We ask about the facilities. Are they secure? Do they have access to hot water or not. Is there soap there? All the same things that are in the national lab certification program are reviewed as part of the inspection of these sites.

Personnel. Who is doing the collection? Are they trained in collection? Is that training documented? Are they aware of bio-safety issues? Many of these officers are very fearful of handling urine samples. It's not in my job description, I don't want to do it. We give them the OSHA blood borne pathogen regulations, my review of those regulations, what they say about urine handling. We make sure they are knowledgeable in that area.

Many times I will go into an office and I'll say who does the collection? Well, we have ten officers do the collections. I will go down the hallway and I'll knock on somebody's door and I'll say, do you perform collections? Come with me. I'll actually have them walk me through a collection, just to see, is every officer as knowledgeable about the collection procedures. I will take the urine specimen. I'll drop it on the floor and I'll say, now what? Now what do you do? Do you have any published policy and procedure that tells you what to do? Are you knowledgeable about these issues?

The custody and control form. Remember, there is no mandate that they use any particular chain of custody or control form. Each site is literally free to develop its own. I make sure that custody and control form comports with the criteria that are established for the SAMHSA program. They have exemplars, for example, of all the SAMHSA laboratories and what those forms say. I will make recommendations about what needs to be on each form. We will actually walk through the procedures. I will say here I am, I'm an offender, I'm here to donate a specimen. I make sure they do proper I.D. on me and we will walk through the whole procedure. If specimens are shipped from a remote site to the testing facility, I make sure that is done under proper chain of custody and that the site actually has some oversight of the remote site. Have you visited that remote site? Do they have secure refrigerators? Do you monitor the temperature? How are the specimens shipped? Do you have chain of custody from the courier's standpoint?

This is the level of rigor that I expose them to. That's collection.

As far as testing goes, we look at the facilities. Again, are they secure. Do they have environmental controls. The test kit manufacturer says this instrument is supposed to be operated between these temperatures. Do you record the temperature? Well, no. What is the defense attorney going to do when he reads that package insert and you say you haven't monitored the temperature? I mentioned that. I teach at the National Judicial College and I spend as much time in the Law Library at Stanford as I do in the Medical School Library. I have probably 1,000 published opinions in my files. I bring up a lot of case law examples to them saying, here's an actual case that occurred. How are you going to defend against that? Again, our goal is to make sure that their results are not only scientifically defensible, but also defensible in court.

Laboratory safety issues, biohazard. Have you contacted your state and local regulatory authorities? Is it okay for you to put these urine specimens in the dumpster? These are the kinds of things I query them about.

Personnel. Who is doing the testing? Virtually all of these testers have gone through manufacturers' training. They are certified. I will truly challenge them about their knowledge of the instrumentation. I will say here's the printout, tell me what this number means. Just to make sure they are really knowledgeable about what they are doing.

We review the analytical procedures. This is sort of the general process by which the specimens are tested. Maybe they are collected over a period of a few days, they are stored in the refrigerator and then they are pulled out. They are allowed to warm to room temperature and then they are processed, and we go through the general issues.

I also developed checklist questions for each particular analyzer. Most of these sites use the ETS bench top analyzer from Dade Behring. A few sites use Hitachi's, they are higher volume sites. That's pretty much the instrumentation. I will ask them questions. Here's what the manufacturer says about operation of this equipment. Are you following all of these things? I make sure they are following those standards.

Instrument maintenance. Just having a maintenance checklist. Specimen shipment. That is if you have a positive result or you are packaging specimens, negative results, and shipping them off to a confirmation lab, how is that being performed? Again, I look at all their documentation, chain of custody, recordkeeping, making sure that is done in an appropriate manner.

Confirmation testing. I insist they have records of any discrepant results. The defense attorney is going to subpoena those records. They are going to say, here we have the report from one of the SAMHSA labs and they said it was negative and you are saying it's positive. No offense to the probation officer, pre-trial service officer who is doing the testing, but who are we going to believe here and does this show sloppy lab procedure? In the same way that a SAMHSA lab may lose their certification, these laboratories may also suffer from having discrepant results. I insist that each one of these sites has a discrepant log where they document review of any discrepancies and make a determination as to why that discrepancy occurred and whether any remedial actions are necessary. We look at how the results are reported, are they secured, do they use any electronic data transmission, et cetera.

Recordkeeping. I might say March 3, I need to see everything that happened in your laboratory on March 3. Pull out the records. I want to know how many specimens tested positive that day, how many tested positive before my client's specimen, this sort of information. I don't really care how they keep their records as long as the records are sufficient for me to do a scientific review.

Finally, their participation in proficiency programs. I review all of the results from the specimens that Bob sends, and if they have had any failures, whether they be positive or negative failures, I make sure they have documented that review.

This gives you an idea of the level of scrutiny that we put these laboratories through. I provide an inspection report, maybe a six page report, where I basically write a paragraph about every one of these sections, and I make a point in that inspection report to note all the good things they are doing, not just the things that need attention. Again, this is going to be subpoenaed as a piece of evidence that they are going to use to show that they know what they are doing. We had Dr. Kadehjian come in and he's inspected our lab and here's what he had to say. It's important for me to have all the good things in here as well.

It takes about three hours for me to do one of these inspections, depending on how much chatting I do and how much case law gets reviewed. It really has proven to be a tremendous educational opportunity for them because they will say, we had a case and this guy said he used Novocain and that's why he's positive for cocaine. Unfortunately, I do get asked that question all too often. It's a chance for me to teach them some of the scientific issues that we deal with. I'm also available on call to all of these sites. If they ever have a question, something of concern, they can call me up. I'll respond to that question. Sometimes they will send me data to assist in the interpretation of results.

We really give these sites a tremendous amount of oversight and we hold them to extremely high standards. I will tell you that I've been extremely impressed by virtually all of these sites. They really do an excellent job. These people are professionals and they care a great deal about doing it right. I've been very impressed with the management of this whole program.

QUESTIONS AND ANSWERS

MR. FORTUNA: This question is for Mr. Baer. You talked about the sweat patch results, about correlating with the urine. Does that vary by the drug type, marijuana versus urine as cocaine versus urine?

MR. BAER: The drug predominately discordant is methamphetamine. Neil Fortner can correct that better, but our experience in our District is that we have more methamphetamine positives in the sweat patch and negative urines.

Perhaps Dr. Kadehjian might speculate on how on-site inspections would be done for districts predominately using non-instrumented devices, since I think our volume in the Central District is as large or larger than any district using instrumented.

DR. KADEHJIAN: Number one, there would clearly be many, many, many more sites. Unless you clone me, obviously lots of people are going to have to be trained to do these inspections. Bob and I have also put in place an inspection checklist which we in fact used Jim Baer as the guinea pig to just try out these questions. What criteria would you ask and how would they be different than the checklist that I showed you, again, absent an LCP criteria for on-site testing.

We do take advantage of the Nuclear Regulatory Commission model, which does allow on-site testing, and I will speak about how we inspect those facilities tomorrow. Without question, doing inspections of many remote sites will be a challenge and I don't think we have an answer for the actual mechanics of how that would be done, but I think it's important and clearly something that is going to be a challenge to figure out how to do.

MS. BOONE (LabCorp): We have about 800 patient service centers. We do have a certification program and an inspection program that takes place on a quarterly basis. I guess I will be talking about that a little bit more tomorrow. It can work. It does work and it's a national program.

MR. FORTUNA: When is the appropriate time for some of these on-site tests to be used or not used, under what scenarios would they be used? The other question I had is there is a discrepancy right now –

DR. WILLETTE: Just to get things started, our experience in the workplace is the greatest interest is in companies that have been doing traditional urine drug testing, I'd say the number one is for employers that have contractors coming onto safety sensitive sites, like chemical plants, refineries and the like, where they are coming in to do either construction, retrofitting, repair work, and typically they are badged to go on the premises. They may require a drug test before they enter, but they would like to know before they are badged and put on the site whether they are positive or negative. I think that is the number one request that we get. We have some international companies that are trying to do pre-employment testing overseas. They want to get rid of the negatives because trying to ship them to the U.S. or to a qualified laboratory to get confirmation testing is logistically very difficult. They just want to deal with the positives.

Occasionally, we have had employers that have gone into big hiring surges. Rather than putting people through an all day physical exam with hearing and all of the occupational health tests, they would like to know before they start that process whether they are dealing with a positive or negative. If they get a positive, they stop the process, get a confirmation, have the person come back if it's not confirmed, and then continue the medical exam.

Random testing and others, I don't see as much use for, but certainly I think you could probably adapt these to almost any, depending on the situation, the scenario, the volume. There are a lot of issues that govern what is the most appropriate way and what is the most cost effective.

MR. FORTUNA: Second question is the differences between the thresholds in the test kits from the --

MS. MURDOCH (Bensinger Dupont & Associates): What we generally advise our clients is that they should consider using on-site testing devices when what they need more than anything else is an immediate result, and when the benefit of obtaining that immediate result outweighs any of the costs that might be associated, and you are going to see it in the pre-employment, where they want to get immediate negatives and put people to work. Occasionally, it's useful in the follow up where you get that benefit, as mentioned by Jim Baer, of the confrontation aspect, if it's positive, or presumptively positive, or the affirmation if it's negative.

I'm less a fan of it in the situations where you have individuals who are already performing some sort of service for the employer because then you have to consider, for example, if you have a device and you are not absolutely certain that you are not going to get false negatives, do you really want to use that when you have somebody performing a safety sensitive function? Is that the current standard of care? You give somebody a non-instrumented drug test. They are negative on the screen, so you keep them in their safety sensitive position, and they are in fact drug using and may cause harm. Whereas, if they went to a laboratory, perhaps they would have been identified as using drugs.

You also have to consider what you are going to do with the person if you get a presumptive positive on a non-instrumented test. Do you stand them down? Do you pay them? Then you have to worry about stigma and disclosing of information inappropriately, defamation, that sort of thing, which is bad.

Post-accident. In my opinion, in a situation where the result has to be really forensically defensible and you don't need an immediate result, post-accident, reasonable suspicion, because you are not going to put the person back to work if they are showing signs of intoxication, even if your test shows negative, I would presume, one would hope. You go to the laboratory and you get the most forensically defensible test possible.

Not that it's necessarily all that much better. It's been well litigated and you are going to have a more defensible outcome. Those are my views.

DR. KADEHJIAN: I would like to make a comment about forensically defensible legal issues, application of these tests, when can we use them, where can we use them. We should remember that we have virtually no case law precedent establishing judicial review of non-instrumented devices. If you look at legal principles, beyond a reasonable doubt, preponderance of the evidence, probable cause, reasonable suspicion, some evidence, those all have percentage figures associated with them. Where do laboratory tests lie? Where do automated immunoassays lie and where do non-instrumented drug test devices lie?

Although we talk about the most forensically defensible result, these non-instrumented devices may be forensically acceptable because they meet the due process requirements for each application, and the due process standards clearly vary for someone who is arrested, for someone who is convicted, for someone who is in a treatment program, for an applicant versus an employee, depending on collective bargaining agreements. All of these due process standards vary for those different applications.

As we talk about devices, we need to consider the demonstrated accuracy that we have seen for these devices, with the studies that Bob and I have conducted, and the challenging specimen sets we looked at, and what the legal community requires, not that we should not try to do the best that we can, but I think we need to remember that those due process standards vary across all of these different contexts. I think that is ultimately what we are going to face.

We need some case law precedent in this area to show how the courts are going to view defensibility of these results, and whether they can be used with confirmation or without confirmation, and what sanctions would apply, so all the things that Julie brought up are absolutely correct. We need the case law to tell us where to go with these. Just consider that as we go through this process of trying to draft regulations for the use of these devices.

DR. OKORODUDU: With respect to the presentations this morning, it was clearly stated -- I just want to give an example, using an example from the emergency room, where the population which you are testing in the emergency room is zero negative or absolutely positive. My concern with the kind of devices at this point is if you apply these devices to pre- employment testing, most of those patients will fall around the cutoff points. If you use a point of care testing device, the result, if positive, will definitely be positive. If it's in the pre-employment or all the examples shown today by the presenters showed that around the cutoff, you will definitely have problems. You will get a lot of false positives and false negatives in your pre-employment setting.

MR. FORTUNA: The next question is the reconciliation of the different cutoffs that the kits use versus the Federal standards, especially for opiates. Are the kits going to be modified or changed to reflect the Federal standards?

DR. CAPLAN: I think what has been presented up to date is that they do meet these requirements or at least most of the kits that are out there advertise that they are targeted at the same HHS cutoffs, the five type drugs.

MR. FORTUNA: Is that true for the opiates?

DR. SALAMONE: The opiates, several manufacturers have already gotten approval for the opiates 2000, and I would suspect that others are in the process of doing that.

MR. FORTUNA: When do you anticipate those being out?

DR. SALAMONE: Like I said, several already have approval and I don't know the stage or the level of the other manufacturers. You will have to speak to each one individually. In general, most of these kits are FDA approved. There are kits that are out there that are not FDA approved and they are used more in the criminal justice area. As long as you see an in vitro diagnostic product, you know it is FDA approved.

MR. FORTUNA: This is a matter of training and maybe a standard for interpreting the results of the on-site kits. I was just curious. Has any testing been done on the light source? Is there a difference if you use natural light or incandescent light, if you use flourescent lighting in the interpretation of these kits?

DR. SALAMONE: That's a very good point. No two eyes see color the same way and we have looked at reading test results under the various light conditions. When something is borderline, it can make a difference between being positive or negative. That's one of the downsides of looking at a color readout because of the variations of people's eyes, their receptors, and also the light differences.

DR. OKORODUDU: My question has to do with quality control, for those individuals using the non-instrumented devices. The NCCLS is looking at single test use and the devices are defined as instruments, particularly the Hitachi 747, it is an instrument. Each of the strips is also considered an instrument. For the purpose of quality control, the Subcommittee on Quality Control for the NCCLS requires that you should take a frequency of use. If you are in a facility with about 15 different individuals doing the testing, if they do testing just once a week, then they are required to do quality control with each test they do, whereas if they do testing every day, then just once a week maybe enough. How do you plan to address this with respect to non-instrumented devices?

DR. KADEHJIAN: You did bring up that the NCCLS does have a committee. Paula Sandtrack is on that. They are specifically looking at non-instrumented devices and how do you QC them. We have yet to see a formal document from them, a draft or even proposed guidance. Currently, these devices, under use in the clinical arena, that is subject to HCFA CLIA requirements, they must now be subject to two levels of control each day of testing. That is the bottom line here. If the tests are used outside of a HCFA CLIA regulated arena, such as corrections, in corrections, there are no requirements for these laboratories. Currently, there is no requirement of how they do quality control, and what could happen, somebody could raise in court, well, you are not doing quality control. They would have an expert get up and say here are the quality control standards, and the defense expert would have to get up and say but, you know, our results are still accurate and reliable. There are no statutory requirements for any controls or lab procedures within the corrections community. Also within workplace testing, although workplace testing is supposed to be regulated under the HCFA CLIA requirements, Secretary of Health and Human Services Sullivan basically put that on hold and it has been on hold ever since. Currently, workplace testing is not regulated under the CLIA criteria. It may be regulated under Federal SAMHSA regulations. It may be regulated under various state laws. For the most part, workplace testing does not have those QC criteria.

I absolutely agree with your concerns about it and I think the industry is clearly concerned about it because they need some guidance on how to recommend to their consumers how to QC them or what level of internal QC is going to be required. It's an area that is unique and I would hope NCCLS acts more rapidly to provide some guidance or some other regulatory agency, such as Division of Workplace Testing.

DR. CAPLAN: We are a little bit ahead of schedule. We will reconvene with the manufacturers and distributors' presentations at 2:00, according to the schedule that is posted there.

MANUFACTURERS AND DISTRIBUTORS PRESENTATIONS

DR. CAPLAN: We are going to start with Dave Evans who is representing NOTA.

PRESENTATION BY NATIONAL ON-SITE TESTING ASSOCIATION

MR. EVANS: It's nice to see that on-site testing is getting the acceptance that it is now getting. I have been involved in on-site testing now as an attorney almost ten years. A couple of years ago, I spoke to DATIA, the Drug and Alcohol Testing Industry Association, which used to be called NACS, the National Association of Collection Sites. At that time, at least my interpretation is that at least a lot of them were opposed to on-site testing. They saw themselves as being connected to laboratories. I got up at their annual conference and made a presentation about on-site testing. In addition to being a lawyer, I am also a sheep farmer. I'm used to stepping in it. I really thought I had stepped in it that day.

Now, things have evolved to the point that DATIA is now meeting with us. This afternoon, we are having a NOTA Board meeting to talk about setting up a joint certification and training program for on-site test operators. It has been the position of NOTA for several years that we want some kind of a certification process. We have built that into the state legislation that we have supported. For those of you who are not familiar with NOTA, let me tell you a little bit about us. First of all, who are we? We are an advocacy and resource center for on-site drug and alcohol testing. NOTA membership consists of the consumers, manufacturers and distributors of on-site tests. We are a trade association. We are also a lobbying association. We are a for profit corporation so we do lobby. Most of our activity has been on the state level getting state legislation that specifically permits on-site testing or getting states to change their position regarding on-site testing.

Our Board has some of the major players in the on-site testing field. We have Biosite Diagnostics, American BioMedica, Roche Diagnostics Corporation, Drug Check, MEDTOX, Casco Standards just got off, but I believe they are getting back on, I hope. LifePoint. We have two other organizations that I believe will be coming on board relatively soon. We have laboratories within our membership. They either are now or have been recently members, Northwest Toxicology, Smith-Kline Beecham, PharmChem, Clinical Reference Laboratory.

It has been gratifying for us to see that some of the larger laboratories have been incorporating on-site testing as part of the package that they present to their customers, because it allows them flexibility, it expands drug testing, and the customers are asking for it.

What I'd like to do is give you an update on where on-site testing is now in the United States. Of course, you are familiar with where we are on the Federal level. We haven't yet been given the okay by HHS, but we are hoping that will eventually come to pass if the regulations are not too burdensome and if the regulations do not destroy the value of on-site testing by incorporating too much expense.

Right now in the United States, 38 states and Washington, D.C. have little or no statutory restriction on on-site testing. Let me just break that down for you. In 21 states and Washington, D.C., there are absolutely no restrictions at all on on-site testing to be used in employment. Ten states permit on-site testing, but only as an initial screening test for pre-employment. Positive tests have to be confirmed by a laboratory. In eight states, there are voluntary employment drug testing laws. These are voluntary laws. For example, Georgia is an example of that, where Georgia provides a Workers' Compensation discount, a premium discount, if you set up a drug free workplace program in accordance with their state law. You don't have to do it. If you want to get the discount, you comply with this law.

A lot of small employers do not bother with this because it's simply not worth it to them to go through all they have to go through to set up a drug free workplace program in accordance with the law. They don't care that much about the discount and they want to utilize on-site testing.

In 13 states, there are some restrictions. Montana requires written procedures that may include the use of a laboratory, although we don't have a firm opinion yet in Montana on that. In seven states, laboratories must be used because a state employment drug testing or laboratory law requires it, and in four states, the clinical or medical laboratory laws, and these are not employment laws, they are clinical or medical laboratory laws which have been interpreted by state officials to cover drug testing in employment. We happen to disagree with those interpretations. We believe that those laws were established for medical diagnostic testing, and as the Americans With Disabilities Act states specifically in the Act itself, a drug test for illegal use of drugs is not a medical examination.

In recent years, NOTA has been involved on the state level. We have gotten either through NOTA and NOTA members and working in partnership with other organizations on-site drug testing laws passed in Oregon, Idaho, Alaska, Iowa amended their law to allow on-site testing, and the most recent ornament in our tree is Hawaii. This year, Hawaii enacted an on-site drug testing law amending their state employment drug testing law that had required exclusive use of laboratories. They now permit on-site testing.

NOTA members were instrumental in getting the Kansas Attorney General to interpret their drug testing laws saying that on-site testing can be used. California has recently gone on-site. Their state clinical laboratory office has re-interpreted their interpretation of their medical laboratory law, saying it does not cover employment drug testing. This has been our interpretation all alone. California has now gone on-site.

Pennsylvania went on-site. Again, a similar situation, approaching the Attorney General there and the state clinical laboratory office said you are right, we are medical diagnostic, we are not employment.

Louisiana, we believe is about to change their opinion on it. There is a bill that the Lieutenant Governor of Tennessee invited us to submit a bill that I believe if it's not already an active bill in Tennessee, it will be, amending their state voluntary employment drug testing law to include on-site testing. This was solicited by the Lieutenant Governor of the state.

New Jersey has an on-site testing bill, a comprehensive drug free workplace bill passed in New Jersey Assembly. This is probably the thing that NOTA worked the most on this year because we wanted to get an industrialized state, populated industrialized state to go on-site. New Jersey is already on-site. You can do on-site testing in New Jersey because the state clinical laboratory law has been interpreted not to include employment. We wanted to get it codified.

It got through our Assembly. It went through our Senate Committee. NOTA testified. We had significant opposition from labor, but we had the State Business and Industry Association on our side, Pfizer, and a bunch of other New Jersey corporations, and it got through the Senate Committee. It got up to the Senate Caucus, and unfortunately, lobbyists are not allowed in there. The sponsor of the bill didn't show up to explain the bill, and a Senator raised an issue about can this drug testing test for Viagra. I guess it made enough of them nervous that they tabled the bill.

We are still working on New Jersey right now. We have to explain to them that the statute specifically says you can only test for illegal drugs, not Viagra. Hopefully, once we get that cleared up, maybe we will bring in Bob Dole to help us. I don't know.

New York. There is a bill in New York. We have been working with the state legislator there. The New York clinical laboratory office has also been engaged in discussions with us about a change in their opinion. We are not sure which way that is going to go at this point.

Florida also has a bill. We worked real hard on Nevada this year. We were not successful in Nevada. We had a bill, Senate Bill 180 in Nevada, that would have changed their state law to permit on-site testing. There was a laboratory that basically got there first with the most, and lined up the casinos against us, and despite our efforts, the bill did not get released from Senate Committee. That is one that we lost, but we still are hopeful and are going to continue to work on it.

We do have a model on-site drug testing bill, an employment bill. It has been distributed to all the state legislatures, and elements of that bill have been incorporated into several of these states that I've been discussing.

If we have another minute or two, I'd like to just go over some of the states and some of the provisions of the bills about on-site testing that will give you an idea of some of our philosophy on this.

Alaska, for example, specifically requires that if you are going to be doing on-site testing, the administrator must be trained, has to receive at least 60 minutes of training on alcohol misuse, 60 minutes on misuse of drugs, and they have to be trained in person by the manufacturer of the test. They have to be trained in proper procedures, evaluation of the on-site test, and they have to be certified that they are, that they are competent to administer the test. They have to be trained to recognize adulteration of the sample and they have to pledge confidentiality.

This is similar. You asked what is the precedent for allowing the manufacturers to conduct the testing training of the test administrators. This is what is done with the Federal DOT for the on-site alcohol test that Federal DOT uses now. I've taken the training. The manufacturer trains you for an hour or two. You see a video tape. You have to go through procedures to show you are competent to administer the test. You take an exam and you become certified. We are modeling it on the successful DOT model for an on-site test that has already been accepted.

Hawaii, it's a very similar process. Here the test can only be used for pre-employment. They have to follow the FDA package insert. Again, trained by the manufacturer and so forth.

NOTA, by the way, only supports on-site tests that have been cleared by the FDA for commercial distribution. We do not support tests that have not been cleared by the FDA. We also only support alcohol tests that are approved by the DOT. We are selective in that regard. We know there are some tests out there that are not very good. They do not have our support.

Oregon, a similar thing. They have to provide a certification to the state that they are administering a test according to the FDA insert and instructions of the manufacturer. Chain of custody procedures have to be followed. They have to be trained in the tests. Any positive result, if there is going to be any action taken against an employee, has to be confirmed.

I was asked just to talk a few minutes about NOTA. We are continuing to look for opportunities to expand on- site testing and what we think is a responsible professional way. We are looking forward to continuing to work with HHS in devising some mechanism to be able to incorporate on-site testing into the Federal drug testing programs.

DR. CAPLAN: The next presentation will be the first of the manufacturers, Dade-Behring, Ken McNeil.

PRESENTATION BY DADE BEHRING

MR. McNEIL: I'm a Marketing Manager for Syva, a Division of Dade Behring. We want to sincerely express our appreciation to the Committee for inviting us to present our products to be used for on-site drug testing. Before I start, I'd like to digress a little bit. This is my first involvement with a committee such as this. I've been involved with drug testing essentially my whole professional career, which is 26 years with Syva. Over those years, I've observed drug testing as improved, as an industry, as a technology. I think it is as a result of committees like this that contribute their time and efforts to doing things the right way, and you should all take, I think, a moment to acknowledge your efforts and see how you have improved things.

What I will present today are our products that Syva markets that can be used for on-site drug testing. Two of them are EMIT based products. The first one is EMIT drugs of abuse assay. It's an enzyme immunoassay, two reagents with separate buffer. That is the one that is used primarily on the ETS analyzer. The second one is the EMIT II product, which is also an enzyme immunoassay, but it's a two reagent, one step prep. It's used for the larger automated systems. Both of these EMIT's can be used on-site with small bench top analyzers.

The second series of products are the non- instrumented products that Syva markets. The first one is the Syva rapid test. This is a solid phase immunoassay product and it's available in configurations of single tests and up to ten test panels. It's manufactured for Syva. The second product is the Syva rapid test. This is once again a solid phase immunoassay, non-instrumented, but it's an integrated sample collection and test system. By that we mean the sample is collected, you can actually test, the chromatographic strips are in the cup, and you see the result in a five minute period. These are available in three to six test panels.

What are the performance characteristics? I will just briefly go through the various assays and their cutoffs. The drugs of abuse assays, the d.a.u.'s, we have a wide range of products. Cutoffs include products that reflect the SAMHSA guidelines as well as other cutoffs.

Likewise, the EMIT II products, there is a wide range of cutoffs. Once again, they do cover the SAMHSA levels as well as other cutoffs.

The Syva rapid test, there are ten different single assays. Once again, the SAMHSA cutoffs are there, as well as the other drugs that we test for.

The Syva rapid cup, we currently have six assays and three more pending FDA approval. The top six are available in three to six test cups.

As far as our on-site products, specimen treatment, there is no special treatment of urine required, that you test urine as it is received in an urine cup. All the products that we commercialize are cleared through the 510(k)'s with the FDA.

Performance summaries. What I have summarized in the next few slides are the data from our package inserts. They are the data submitted to FDA to demonstrate efficacy of the product.

As I work through it, you will find that the use of reference systems might have changed. I think part of that is due to timing. When we go back to the d.a.u. products, a lot of those filings were made in the early 1980s, so the reference might have been to a TLC, but not as diligently against GC/MS as we do today. For the d.a.u.'s, the predicate device in most cases, with the exception of cannabinoids, was the previous generation of EMIT that was available at the time. The reason for that is once again, FDA expected you to show that the product was substantially equivalent to an approved product, and EMIT was the only one there at the time. That was the predicate. We went to confirmation with either GC/MS or TLC, in the case of amphetamines. The cannabinoids were done against GC/MS solely. For the other d.a.u.'s, once again, the predicate was either MS or in the case of methaqualone, GLC/UV method, and went to confirmation, GC/MS in most cases.

EMIT II, these were introduced in the early 1990s. MS was at that point the established confirmation method and the positives that were run during the clinical trials were sent to GS/MS, and that shows the confirmation. Likewise for the rest of the EMIT II products, the predicate would be the EMIT, or in the case of LSD, an RIA, and then GS/MS was used to confirm all positives.

For the rapid test, the manufacturer supplied and did the clinical trials on this, and once again, the predicate in all cases, except tricyclides, was the EMIT assay and then MS was used for most of the confirmed positives.

For the Syva rapid cup, once again the manufacturer generated the clinical data and they used a mixture of another EMIT and GS/MS.

As far as how we propose dealing with adulteration at this point in time, as far as on-site, we recommend the use of the Delta Check IV, which is a dip tip that can check for four different adulterants. If an instrument approach is desired, then the UA perfect reagents are available from CR&C.

As far as how to evaluate prior to use, I think it is important for the labs to understand which drugs they want to monitor, what the cutoffs are they need to monitor, what regulations they want to comply with.

The volume of testing is crucial. You want to match the device to your volumes. What your available staff levels are and confirmation approach to be applied.

As far as unique features of the instrumented on-site, with EMIT, there are a wide variety of instruments that can be used. We have a multitude of applications. The reagent costs can be relatively low. Results are less subjective in that you get an instrument reading, and then finally, there is a higher throughput on the instrumented systems.

As far as the non-instrumented tests, the features are once again ease of use, which has been a resounding theme here, that operators do not need as much training sophistication to run the non-instrumented test as the instruments. There are built in controls for these devices, and there are a wide variety of product configurations. The SRT is available in singles, up to ten test panels, and the SRC is currently available in three to six test panels.

Stability. You only use the device when you are ready to test, as opposed to a liquid reagent when you open it and the reagents may degrade.

Control recommendations, another question we were asked to address. For the EMIT d.a.u., which is the oldest configuration, the control recommendations are that daily calibration be validated by assay and positive and negative controls, and then if results fall within acceptable limits, as the lab defines and calibrators meet the separations, the minimum steps that are placed on our base labels, then run samples.

With EMIT II, the recommendation is that positive and negative controls be included in each run. Once again, if they fall within the acceptable limits as defined by the laboratory, run samples, and then also in many cases, we provide expected calibrator separations, which gives a lab an idea on whether their response curve meets criteria.

The rapid test, once again, it's a single use device. Internal controls, there's a negative and positive process control built into the device, and also external controls are recommended to be tested at regular intervals.

Rapid cup, similar to the SRT, has internal controls built in and to be tested at regular intervals is a good laboratory testing process.

As far as external performance testing, commercial controls are available and we recommend using other manufacturers' controls. This is maybe true for everything except LSD, which has been kind of challenging.

Proficiency testing, to participate in proficiency surveys to learn how your lab is performing relative to other labs in stressing the system a little bit.

Training recommendations. For the instrumented on-site test, we recommend that the key operators be trained by Dade Behring technical staff, and that can be done at the laboratory site. Non-instrumented, it's a more simple orientation and training can be done by Dade Behring staff or others.

Use of the devices. The instrumented tests, we recommend for use in the higher volume areas and the non- instrumented in low volume and when turnaround is an important factor in the test.

That concludes my presentation. Thank you.

DR. WALSH: I don't know if you want to discuss all of this but one of the things that Yale said is it would help all of us, and I wondered, is there any technical difference in the product that you sell and the product that PVM sells under their own name?

MR. McNEIL: Not to my knowledge. It's my understanding it's the same product and it's labeled differently.

DR. CAPLAN: What was the company? Princeton.

MR. McNEIL: One question was who manufacturers. The EMIT's are manufactured by Syva. The SRT is manufactured for us. SRC currently is manufactured for us but we are in the process of bringing that manufacturing in-house.

DR. SAMPLE: The question I have and really this question I'm going to ask of all the presenters today, so if you could care to answer this so I don't have to get up and ask the question each time, I'd appreciate it. My question relates to the separation with the cutoff differentiation at plus or minus 25%. Is that something that you feel your product is capable of?

MR. McNEIL: As far as the performance around the cutoff, the d.a.u. products and in particular the EMIT II products, were optimized to meet that standard, and they perform very well. The non-instrumented approaches, we are still in the process of evaluating how well it meets it, but in some data, we have seen that it performs acceptably but in other cases, we have seen issues making the plus or minus 25%, and the test points might be as far as plus or minus 50%. The instrumented, yes, plus or minus 25%. The non-instrumented, right now the jury is out in my opinion as far as whether we will be able to support that as a claim or not.

DR. CAPLAN: Is there a percentage you might support, if it's not plus or minus 25%, is it plus or minus 50%?

MR. McNEIL: At this point, we are testing at plus or minus 50%. Obviously, there is a desire to make plus or minus 25%, but there are data that shows the devices have done that, but we are not showing it routinely.

DR. CAPLAN: We will move onto the next presentation, which is Dr. Lappe with National Medical Review Offices.

PRESENTATION BY NATIONAL MEDICAL REVIEW OFFICES

DR. LAPPE: The product that I'm going to show you today is a product that we have developed through our experience with drug testing in the workplace. As the product continues to be developed, we will see some of the applications that we don't currently participate in, like criminal justice and other things, but this is again a product that is designed for workplace testing. It is what we call a point of collection screening device to identify negatives or what we are going to call perfect negatives, those negatives that don't contain adulterants, and we will define that in just a little bit.

It is not just a product but a system made up of a number of products, which we will describe. It includes some services and some information technology that ties it altogether.

In coming to meetings like this, one of the things that I really like are buzz words. I'm going to give you some new ones. "On-site," we never really liked because when we started using that ten years ago, we thought drug testing was going to make its way to the point of hire. At this point, as long as urine is the matrix, it's not likely to migrate to the extreme left, except in those workplaces that wish to collect urine samples. For the most part, urine based on-site will live at the point of collection.

This is not a non-instrumented drug test, although I really like that term better than "on-site," but it is a hybrid really of a monoclonal assay in a cup and a reader that interprets the result of the strips. To give you the whole thing, it's a solid phase laminar flow monoclonal assay strip with adulteration and integrity strips in the cup, and that cup is self aliquoting. We will talk about how it does that. The point is collections are performed in the normal way, the way that they are performed in a DOT mandated drug screen or non-mandated drug screen. This screen is actually performed under seal with a self aliquoting mechanism.

Once the collection is done, it's put into a reader, into an instrument, which collects the data using a scanning system, and reads the bar code information off the cup. It reads the test panel codes off the cup. It scans the lines on the assay strips, reads the color changes in the adulteration and integrity strips, and interprets them. It produces a perfect negative result. "Perfect negative" meaning no drugs detected, and none of the other variables failed, like adulteration and controls, and then transmits that result to the host. The strips are FDA approved. The instrument is pending FDA approval.

To dissect the cup a little bit, we like to refer to the cup as a dumb cup and a smart lid. Dumb cup in that it doesn't have any test strips in it and the donor can take the cup to void into without worrying about tampering with the actual test. It's an unique shape because it fits into the reader a certain way. It's collected in the usual fashion, the way a DOT sample is collected. The temperature strip is on the cup. When they come back, the lid, which is packed separately, contains a tamper evident seal already applied to the lid portion. It is bar coded and there are bar codes are the B bottle seal. You can do a split specimen collection using this, and then this portion goes down onto the cup. Once the sample is collected, it's sealed in a chamber, called aliquot chamber, in the bottom of the cup. That aliquot chamber at the bottom of the plunger is the aliquot that is forced up to the test strips. Once the sample is locked in here, the tamper evident tape is pulled down, the donor signs the chain of custody document, and can actually leave the test site. The cup is then placed into the reader.

The reader then plunges the cup and activates the test. There are a couple of things that it does. It starts a clock so that it knows exactly when to start reading and it knows when the window has been exceeded and can no longer read a result. That window is eight minutes. After eight minutes, if it can't produce a perfect negative result, it times out and it says "send this to the laboratory." The other thing that it does is locks the sample into the reader. When the cup is placed in the reader, the plunger is activated. The clock begins. The sample can't be removed from the reader until it has completed its testing process.

I want to go back a little bit to sort of the design of the system and what we were trying to accomplish. If you look at the way drug testing is done today in the workplace, you find that the employer or the end user sends an employee to a collection site where the sample is collected. That sample is then shipped to a laboratory and the result is then shipped to the MRO. In a laboratory centric model, the lab, which is this portion, is in the center of the collection and the MRO process, and we refer to that as the analytical portion, everything to the left of the lab is the pre-analytical portion, and the right side is the post-analytical portion.

The process of drug testing is an analog process until the laboratory has completed the test. At that point, it becomes a digital process. This red asterisk represents the conversion point from analog to digital. The goal in our on-site or point of care, point of collection test, is to move that point as close to the collection process as possible, and to convert the analog information inside the cup to a digital piece of information that we can acquire earlier in the process, but do that without destroying the process.

Each phase of this process is there for a reason, and we don't want the collector messing around with the analytical portion or the review process, so we wanted to maintain the segregation of functions without adding liability to the process by breaching confidentiality and some of the other things that could happen.

The current system, laboratory based or on-site tests, are designed to test for the presence of drugs. We designed this system specifically to test for the absence of drugs and adulterants, producing either one of two results. It's either a perfect negative result or it is indeterminate.

The definition of a perfect negative result is it must meet all the following criteria, no drugs detectable at threshold, creatinine levels exceeding 20, nitrite levels exceeding 500, pH within range, and temperature in range, and it has an external QC, and that the internal controls, each assay strip has its own control, and it has a daily PT, which we are going to talk about. Each day a performance test is performed before the negative samples can be thrown away and before the indeterminate samples are called into the laboratory.

An indeterminate result is any one of these. Drugs are detected at or above the threshold. The adulteration or integrity tests are out of range. The system fails for any reason. The controls fail for any reason. The assay strips' control aren't activated. The bar codes mismatch or are unreadable. The daily PT fails. The external Q/C specimen, which is host derived, or can be host derived, is called in, or the system just defaults to indeterminate.

Anything outside of a perfect negative will cause the sample to be considered indeterminate. Hopefully, the operator won't know, except maybe in a system failure, why the sample is being called in. All the indeterminates are pooled together to disguise the presumptive positives. Presence of adulterants, currently, the E screen system test for nitrite, pH, and creatinine, and with the cup as designed, again, the collection procedure mimics the Federal collection procedures.

The only difference is that each specimen cup is bar coded, and as a result, we can create a chain of custody or custody control form on the fly, originating with the unique cup number that is already applied. You can either use a paper custody control form or the system will print one for you when you put the cup into the reader. Actually, when you scan it before you put it into the reader, using a bar code scanner, it will produce the chain. It will print it on the fly.

The analytical procedure is replaced with the reader and once the collection is complete, the tamper evident tape is pulled down over the cup. The donor has signed the chain of custody form and they can now leave the collection site. The cup is ready to be tested. The aliquot is already locked into the chamber. It's placed into the reader and a controller, which is the touch screen controller, begins the process of plunging the cup and starting the clock.

At two minutes, the scanning begins and it will continue every 30 seconds until eight minutes. At the point where it captures all of the appropriate negative information, it stops. If it doesn't, it will go for eight minutes and then time out and produce an indeterminate result.

Results are recorded by the host system in real time and then the specimen is unlocked. The results are then transmitted to the MRO. At this point, the analytical process has ended and that specimen is set aside until the end of the day. At the end of the day, the operator runs a shipping log to do two things. One is to run a PT for the day, and the second is to find out which samples are to be discarded and which ones are to be sent to the lab.

On the post-analytical side, all interpretation is by the reader, both the translation and the transcription of the test strips are performed at the level of the reader.

The post-analytical procedure doesn't really change. The host system reports both the negatives and the positives to the MRO as if it were a laboratory.

Daily performance test, again, what happens when the end of the shift occurs or the end of the day, the operator runs a daily shipping log from the E screen controller. Immediately prior to that shipping log printing, the controller says you must put in a red or blue cup, which they have on-site. There are 12 variations currently in the red cup and that's an indeterminate cup, and blue cups are perfect negatives.

If the system says to put a blue cup in the reader, the operator puts it in, hits the "enter" button, it scans it, if it works okay, it releases it and prints the shipping log. If it doesn't, it says to send all these samples into the laboratory.

Going back to the test data, we started with the strips alone, read visually, using positives confirmed by GC/MS to determine the false negative rate. Those are broken down by drug. You can see that each and every time it was able to detect the presence of drugs visually read.

The next study was done using EMIT for actual samples. I think there were 309 samples in this study, comparing visually read strips to Hitachi 717, and these strips had zero false negatives, two false positives. The false positives had drugs present below the threshold.

The third study, there were 415 human samples from a laboratory. The test strips were read electronically with the reader. Each sample was tested by EMIT for all five drugs. Actually, this was about 2,000 and some tests actually performed. 315 were negative by both the EMIT and the reader; 52 were positive by both EMIT and the reader. They all confirmed positive by GC/MS. 35 were positive by the reader and negative by EMIT; 18 of those were PCP positives, where we discovered a shadowing error in the cup, which was later corrected. Of the 17 remaining positives, 13 had drugs present below the cutoff, three confirmed positive by GC/MS, two for THC, one for cocaine, which demonstrated the three false negatives EMIT's in the study. 13 were positive by EMIT and negative by the reader, and were GC/MS negative.

Some more unique features of the system are that the test is not human readable. When this test is performed, you cannot read it. It is encoded using bar code technology and it is decoded electronically. Therefore, there is no human translation of the result, meaning no human interpretation of the result, and no transcription of the result to a piece of paper or to a person or to a system. It is all captured electronically and transmitted or transcribed electronically.

There is no operator bias of the result, since the operator has no responsibility for the result. There is no visual acuity issues. The reader will not read outside of the window, the two to eight minute window, which is recommended for reading these strips. The results are maintained in confidence and it is not until the end of the day when someone has to decide which samples to ship to the lab that there is any knowledge that a sample may not be negative, although there are external Q/C's that are called in and pooled with negative results.

Additional benefits are there are some software that drives the thing, which prevents an operator or the collector from failing to complete a portion of the chain or part of the process.

We are hoping in the future to be able to tune the cutoff levels using the reader to get within the current capability of the plus or minus 50% into the plus or minus 25% range.

MR. WILLIAMS (Doctors lab): When you talk about built in internal QC, what is this internal QC? It obviously is not positive samples and negative samples.

DR. LAPPE: There are internal controls on the strips themselves, so there are strip controls. The reader goes through a self calibration type of control before each test.

MR. LINEWEBER (UTU): What is the time frame of the communication between this and the MRO?

DR. LAPPE: It depends on who the MRO is. This is an Internet appliance. It's immediate. The server is picking up the result in real time. If the result is read in three minutes, the server has the result in three minutes. If the MRO is linked to the system electronically, they will have it within 15 minutes. If that sample goes to the lab, they are going to require the result from the lab.

MR. LINEWEBER: The server will send that to the MRO, so the MRO would know then to withdraw that person from any safety sensitive function, pending the laboratory analysis; right?

DR. LAPPE: That's a company policy issue, if that's what the company has instructed the MRO to tell them. Right now, there is no mechanism for that, for knowing which samples -- in a laboratory centric model, you don't know what samples are pending that have failed the EMIT and are pending confirmation. When you separate the screen from the confirmation, you now have access to that knowledge, if you can get it. The obvious one is when you are doing the test right in front of you, you know which ones are negative and which ones are going onto confirmation. The regulators will decide what to do with that information in mandated testing, and for non-mandated testing, the company policies will dictate whether or not to pull somebody out of work, let's say a safety sensitive job, if their sample is pending. There are some liability issues.

DR. FOLEY (LifePoint): Your answer on internal quality control wasn't clear to me, but over and above that, I'd like to know how you calibrate, how stable is your calibration, and at what wavelength do you read your strips? In any order.

DR. LAPPE: The exact wavelength is a green wavelength. The reader has a shield and it blocks out ambient light and uses its own light source. The self calibrating issue, I don't think I can answer because it is a software issue, it's a software/hardware issue that was developed by somebody else. I probably can't answer that.

MR. PATTEN: What about stability of calibration?

DR. LAPPE: I'm not sure we can answer it entirely at this point, but it is designed around a software model where every day, the PT is looked at and it is looking for a standard at a certain wavelength, and if it's off, we can tune the reader centrally, so it can be done remotely every day and it has to be done. The actual parameters around the model, I don't remember what they are right now. It's a dynamic model that every day will go through as the system is turned on, before the first test is done every morning.

DR. SAMPLE: There were a couple of items on the initial PowerPoint slides that were shown this morning that was asked of the manufacturers that I'm not sure I heard answered, and perhaps you could just touch on those briefly. One, what are your recommendations for quality control in terms of actually running urine samples that are control samples? Do you have recommendations with respect to external PT and training on use of the device?

DR. LAPPE: The first was external QC's?

DR. SAMPLE: Urine samples that you run as a control.

DR. LAPPE: The way that it is set up today is that the system calls in samples after they are complete. "Calls in" meaning it randomly calls in 3% of the samples to be sent to the laboratory for EMIT screening and confirmation and compares that to the result that the reader determined.

The second one was training?

DR. SAMPLE: Training on use of the device; yes.

DR. LAPPE: Currently, the training is going to be done during the installation process. There is on screen training and on screen help that is available, and there is tech support. The system is protocol driven. There is not too much that the operator can deviate from. The screen is telling the operator to scan the bar code, to check the temperature of the sample, to check for any signs of adulteration, the same things that are currently done without protocol reminders under Federal collections, and will not move forward until the operator has answered those questions.

DR. BUSH: I have a list here and I have to fill the list out for everyone so that I'm happy with my data. How is the device evaluated prior to use in a work site?

DR. LAPPE: How is each instrument set up or calibrated?

DR. BUSH: A daily calibration. How is it calibrated and set up for use every day?

DR. LAPPE: The daily calibration occurs at the end of the day. It's timed to the shipping log so that samples aren't discarded if the external PT fails for any reason. We do the calibration. If the calibration is okay, it then says you can trust the results that you got since the last PT, therefore you can discard these negative samples. Keith tells me that when it's turned off and turned back on in the morning, it goes through a calibration check.

DR. WALSH (Walsh Group): My question is if at the end of the day, you get a bad QC and employers have been downloading negative results all day and now you are sending things onto the lab, it seems to me you are going to get yourself in a mess down the road.

DR. LAPPE: That could be. I have no answer to that.

DR. CAPLAN: Okay. We will accept that answer, I guess.

DR. LAPPE: Thanks.

PRESENTATION BY PHARMCHEM

MR. FORTNER: I am going to go discuss the on-site drug testing products offered by PharmChem. They are called Pharm Screen.

Pharm Screen is offered in a variety of configurations. There is a single test. There is a dual test. There is a four test. All these entail the use of a pipette to deliver urine sample into this testing well, and then there is the five card test, which Dr. Willette earlier in the career referred to as "the utter test" for the strips that stick out from it. At any rate, the testing principle behind this, it utilizes immunoassay, monoclonal, solid phase, utilizing colloidal gold as a visualization region.

There has been a number of questions regarding a control line. In the vast majority of these non-instrumental testing devices that utilize a control line, it's important to realize that the control line is not immunoassay driven. All it does is detect -- I was going to say urine, but that's not necessarily true -- a liquid was deposited onto the test strip and it adequately wicked across. That has some implications later, particularly as you try to address potential adulteration. You can have a sample that has adulterants in it, and you will get a positive control line. Will you get a valid test line? It depends on what the adulterant is.

In this particular line of products, a negative test produces a line in the T or the test section. Of course, you have to have a valid line in the control session to consider it a valid application. A non-negative test will produce a blank space in the test line. You should see a control line, positive, a blank in the test line for what we would call non-negative.

The next question asked, what are the performance characteristics of the device. We will talk a little bit more about some of the unique features, but suffice it to say that it uses the SAMHSA cutoffs with the exception, or the addition, perhaps, I should say, originally it was a 300 and back in December when SAMHSA rolled over to the 2,000 opiate, came out with another product line that utilizes 2,000 opiate cutoff as well. Utilizing these cutoffs, the product was evaluated and submitted through FDA using spike controls, 50% above, 50% below the administrative cutoffs. Certainly, at 50% below, we can achieve a negative test. At 50% above, we certainly can achieve a positive test, and in fact, one of the characteristics of this device is it has high negative predictability, which also means it will pick up 25% above.

This device is a little bit skewed, and I'm sure all of you have committed to memory Dr. Willette's graphs with all the different colors and picked out Pharm Screen in there and looked at it. If you go back and look at it, you will see that it has very few false positive readings. In fact, almost none at all. That little green line is very small. It has higher what you would call false positives, because it does pick up some of those that are below the cutoff, but when you get to 50% below, you sort of get the drop off.

Issue three is a fairly straightforward question. Is any special handling required for the specimen? No. This is a urine sample. As long as you are doing proper forensic collection procedures, there is no special handling required. Certainly, good laboratory practices are strongly recommended.

The product is cleared by the FDA. We have had this product since late 1995. In that documentation that we also submitted, and I won't really take the time to go through it because it's a pretty thick file, but there was extensive testing, certainly correlation, clinical evaluations between samples out of our SAMHSA laboratory relative to the on-site performance of this. That information is also contained in the filings to demonstrate efficacy of the product.

How do you recommend dealing with adulterants? That was the next question. Even those clients of ours that do direct observation have some difficulty dealing with adulterants. There is direct observation and there is really direct observation. Even the real direct observation can't address sometimes the urinator or the self catheterization.

We currently recommend that our clients use the Delta Check IV. It does the standard test right now, creatinine, nitrites, glutaraldehyde, pH. This is the more critical aspect. Abnormal or invalid test results or suspect specimens, it's important to realize, and certainly Research Triangle and HHS have been quite involved with the adulteration aspect, that there are many more adulterants out there than just nitrites and glutaraldehyde. Most of them are oxidizers. They will give some degree of reaction, in particularly the nitrite test strip area, and that's why we recommend abnormal or invalid test results be forwarded to the laboratory. Certainly anything that is suspect, samples that have objects floating in them, suspect, I would recommend you send those to the laboratory for subsequent analysis.

Issue seven asked basically how would you recommend that the device be evaluated, and that is both in general and for batch or daily use. Certainly, you want to verify lot number, expiration date, all that good stuff, to ensure you have product reliability, and one of the things that I also want to add into there is that the devices need to be stored properly. They don't have to be refrigerated. You don't want to sit them in the trunk of your car and subject them to high temperatures. You can then destroy the antibody coupling and result in false negative results.

We do recommend the use of both a positive and negative quality control pool. We currently have those spiked at plus/minus 50%. Certainly, minimally, each day the device is used, if you are doing larger batch numbers, certainly incorporation of additional quality control pools at the discretion of the operators would be advisable.

Unique features of the product. As I told you before, it has a high negative predictability, so that when you get a negative result, you have a very high degree of confidence that it is in fact negative.

PharmChem has an unique program with Pharm Screen that uses a unique chain of custody form, specifically designed for our non-instrumented testing device. It certainly helps to establish forensic defensibility from collection through that test through confirmation.

One of our workplace clients is implementing this in their non-regulated side. I think Jim Baer talked earlier that on the probation side, there is a big issue of trying to collect this data. You have different operators that are running it.

In the integrated program that we have here, the collection site can fax Pharm Screen chain of custody to the laboratory, and we have a system that enables us to directly read that information off that chain, all the information for patient data, demographics, results, which then can go into the statistical database. It can feed to the MRO. It can feed to the end result reporting, depending on whether it's regulated or non-regulated. I think you have an ability to help track that statistical information.

One of the other issues, it asked about manufacturing. I think it went into the second to last question, OEM specifications. We developed these specifications a number of years ago. The product was intentionally developed originally for the criminal justice system, as we hold the contract for the Administrative Office of the U.S. Courts. They are more interested in a true negative as opposed to something that may be in there that needs additional confirmation.

Given the issues they deal with, that's an acceptable -- I guess it's a bias, if you will, in the product. It does have a tendency to pick up samples certainly below the plus 25, and even down to minus 25 in that arena.

Quality control. I said see issue seven, the same thing we talked about before. If you have short term memory problems, you can't remember what issue seven was, verify the lot number, run positive/negative controls, at least minimally each day.

One of the things that was asked about, and actually I think Jim Baer brought it up, in looking at correlation of his on-site testing and positive rates, we are subject under the U.S. Federal Court program to blind quality control. They send us 350/450 samples a month, something along that line. Our current rate is 99.3 accuracy with no false positives.

To answer that indirect question of are your positive rates influenced by a laboratory, I don't think from that perspective. Maybe influenced by other things, certainly education and deterrents in the criminal justice arena as you are testing those people is the desired effect and the desired goal to see a decreased positive rate.

We do support external proficiency testing. We currently don't have a program in use but one that would model the HHS NLCP program would be desirable, and I think it would follow.

We currently have a PT testing program in-house for sweat testing. To develop one for the on-site testing, at least from the performance testing aspects, I don't think it would be that difficult and easy to achieve.

Issue 11 under my numbering, what are the recommendations for training. We have a couple of devices, tools, that are available. We have a very detailed training manual, step by step, in terms of how to use the device, how to read the device. It's in black and white. If you want the color version, you have to use the videotape, but there is a videotape available as well. I did supply copies of those to DTAB for their review at their discretion.

How do you recommend using the device in workplace testing settings, controls, et cetera? Certainly, any forensically and scientifically defensible program needs to include proper collection procedures, kits specifically developed for on-site testing, which we happen to have with Pharm Screen. Adequate training and certification in the device, proper reading, timing of the device. Negatives will be available in three minutes, positives in eight. Within the three to eight window is the recommended timing.

Storage and handling of the device. We talked about you don't want to sit it out in your trunk, subject it to high temperatures. Winter is coming here soon. On the West Coast, we will get rain. Here, you will get cold temperature. You don't want to let these things freeze either. We certainly advise the use of positive and negative controls, and depending upon the program, sending a percentage of the screen negatives and all non-negative samples to a certified laboratory.

The last ones that got added. Do you manufacture the product? Well, no, we don't. It is manufactured for us specifically under the trade name "Pharm Screen," under OEM specifications by Applied Biotech in Southern California.

Do you distribute another manufacturer's product? No. Pharm Screen is not provided to other companies for resale or re-labeling. It is sold to distributors exclusively under the Pharm Screen registered logo.

DR. SAMPLE: For the performance characteristics and we are talking about the HHS cutoffs, the usual analytes, I notice that amphetamine was not mentioned, that you are only doing methamphetamine. Is there any cross reactivity with amphetamine, and if not, would you like to comment on that?

MR. FORTNER: There certainly is cross reactivity with amphetamine. I don't remember the exact level. I could pull it out of the insert. It's not at 1,000. I think it's probably around 200 to 300, if I remember. I'd have to go back and look at it specifically.

DR. WALSH: Do you know if this same identical product manufactured by Applied Biotech is sold under different names around the world?

MR. FORTNER: I believe the answer is yes to that question. We have our own OEM specs, so we get lots and they get produced. I have seen what looks to be the identical product coming from overseas. I had the opportunity to evaluate some of those, and they don't meet the same performance characteristics. I don't know if this is material that gets manufactured, it doesn't meet our specs so it gets resold here, there, somewhere else.

MR. FORTUNA: The quality control pool, is that provided to the on-site test collecting?

MR. FORTNER: We have provided that. It's not something we are currently manufacturing in large numbers. We have some clients that have specifically asked for it and we do provide it to them.

MR. FORTUNA: What drugs do they include?

MR. FORTNER: However they want. We have single drugs available or we can do all five. We have our own rather extensive quality control/quality assurance department.

MR. FORTUNA: As a collection site, if we use a quality control and we have a lot of these things and we get a bad test, what do we do? Is it guaranteed then? Do we send it back to you?

MR. FORTNER: We test every lot that we get in and we keep records and we actually sequester parts of those lots. If there is an issue on performance, we ask you to send the material back in and we will re-send you a new supply of it, so we can do our internal investigation on that particular lot or group of product for evaluation.

DR. KADEHJIAN: Neil, you made a comment that your internal control was in fact a process control, that it merely reflected wicking of the solution. I'd just like to point out that the issue of QC is clearly a major one, certainly CAC is concerned about how you establish QC criteria for these devices. One of their comments is that we want to make sure that QC represents reactivity and not process. I should point out that some manufacturers, I can't speak for all, but I know some manufacturers do have devices where their quality control window is not simply a wicking, but in fact represents an immunologic reaction, that there is an antibody/antigen interaction, and they can then argue that QC represents a check of the propriety of the immunochemistry that's going on.

MR. FORTNER: I think that's a very valid point. I didn't mean to imply that everybody that is out there that has a control line doesn't have an immuno reactive. I think it's an issue that the Drug Testing Advisory Board needs to be aware of in the sense that a control line is a control line is a control line, what is it controlling. You could go dip this thing in coffee if you want and get a positive control line. It is just something that the Board needs to be aware of.

DR. SALAMONE: Your control line is still immunological, isn't it?

MR. FORTNER: No. It's not an immunological control, unless the coffee I'm drinking has something in it that I don't know about.

DR. SALAMONE: How is the blue color captured?

MR. FORTNER: It's a red color.

DR. SALAMONE: How is the red color captured on the line? There must be some immunological, non-specific antibody there or something.

MR. FORTNER: It's a non-specific but it's color development release, but it's not immunological in the sense that you are looking for competition of benzoylecgonine or methamphetamine or anything else like that.

DR. SALAMONE: This is important. All of the control lines in immuno-chromatography are immunological. They are not related to the drug at all. There is some other antigen on there that's non-specific, but clearly, your test as well as many of the immunological tests that are based on immuno-chromatography show that there is some antibody reaction occurring. It's just not a specific one.

MR. FORTNER: In our perspective, it shows that you got adequate wicking. Beyond that, I don't know that you can make a whole lot of other conclusions from it.

MS. MASTERSON: I'd just like to make a clarification point. Under clear requirements, QC for devices, especially simple devices that are CLIA waived, are different than those for standard laboratory, moderately complex requirements. I would like to make this point because when you have very simple and easy to use devices that are inherently self controlled under certain conditions, the requirements are only for once a week. That's because there is a certain assumption that the manufacturer has built and standardized their products and that the other internal controls are testing the things that would normally cause a product to not perform. It's the moderately complex reagents that are run daily.

MR. FORTNER: I appreciate that. Our recommendation for a daily control really comes from our experience and involvement from a forensic urine drug testing lab, which is what we are here to talk about before the Drug Testing Advisory Board. That's simply our recommendation. Whether it falls or doesn't fall under CLIA or HCFA I think is outside the scope, of at least this group. That's a good point.

PRESENTATION BY ROCHE DIAGNOSTICS

DR. SALAMONE: Before I get into TesTcups, just to give you a bit of a historical perspective, Roche was really the first company to launch a non-instrument based immunological assay for drug of abuse, and that was back in 1974. It was the morphine agglutex reagent. I don't know if many people here remember that. It was very popular in Asia. Actually, we took it off the market in 1991, so it was on the market for quite a big time. It wasn't such a big selling reagent.

Since then, we developed the OnTrack assay, which was a latex agglutination assay, and that was launched in 1989. While the results of it were quite reliable, there were market forces that were pointing more towards a more simplified test. This involved three reagents, stirring, and a three minute wait as it went up to the agglutination slide.

Back in the early 1990's, we started to ask customers what they wanted. Customers didn't want to handle urine. They didn't want to pipette urine. They didn't want to deal with dropping reagents or stirring. They wanted no refrigeration and they wanted a device that would evaluate multi-analytes at once. Essentially, they wanted to look at somebody and see whether they were on drugs or not.

Trying to put all these characteristics together, we developed the TesTcup. The TesTcup, as is all of the tox assays for Roche, is FDA approved. The cutoffs that we are using for the five panel are the SAMHSA cutoffs. We also have some other cutoffs used for non-SAMHSA compounds, but I won't talk about them right now.

The assay itself is quite simple. You just void into the cup, place the top on it, tilt it for ten seconds, and then you wait for the test valid lines to appear. I'll talk about the test valid lines in a bit. Once the test valid lines appear, which takes about three minutes, you peel off the label on the top and the absence of color means a positive test. In this case, the middle track there is THC and cannabinoids, and that's positive and it's negative for cocaine and morphine. At that stage, if it's positive -- this is just a close up of the test -- you turn the top the rest of the way and it puts it into a locking position, and you can ship it out to a lab for GC/MS confirmation.

Let me talk about the inner workings of the cup a bit. This is a view of the cup looking down towards the center. This here is a sample reservoir that is covered by a diaphragm. What happens here is when you turn the cup, the lid of the cup, the diaphragm is open, and then when you tilt it, 300 microliters of urine fit into a sample reservoir right at the top here. When you stand it back up, only the 300 microliters of urine come into contact with the reaction strips. The rest of the urine is retained. We did this for several purposes. One, to isolate the bulk urine from the reagents, and two, to keep this diaphragm on there so that adulteration would be minimal, in terms of someone wanted to adulterate the reagents. Taking the cup and stripping away the front cover, what you have here are four strips. You have a sample strip, which contains buffers that stabilize the urine and filter the urine. You have a wicking strip here. You have the immuno-chromatographic strip here, and then on the top of that immuno-chromatographic strip, you have a strip that contains the blue colored microparticles.

The way the test works. The blue colored latex particles contain specific antibody. They are located on this top strip. Then you have a detection zone that contains the specific drug conjugated to BSA or some type of protein, and that's immobilized at the detection zone. You also have a test valid zone which contains antibody that is anti-protein antibody, and the specific protein is BSA, and there is also BSA on the microparticle surface, so this is the control immunological reaction band. Then you have the wicking pad.

What happens here is when urine comes flowing through, it is again filtered through the sample pad and starts flowing down this immuno-chromatographic strip, which is made of nitrocellulose. The main stream goes straight down, but some of it splits off and reacts with the top pad, and slowly, the blue colored latex antibody is leeched off. It dissolves slowly.

This is basically a modified immuno- chromatographic design, and the reason why we put this extra pad here is so that more urine comes into contact with the latex particles to increase the sensitivity of the assay. The more urine that can react with the microparticles up front, the more sensitive your assay can be.

This is just a cartoon of it. Again, this is the dried strip, blue colored antibody microparticles, the detection zone and test valid band.

If you add the sample and the sample contains free drug, the free drug finds the antibodies on the microparticles and as they pass through here, there are no antibodies available to bind the drug, immobilized drug on the strip, so all of the blue colored microparticles pass through and land up in your test valid band.

The bottom picture is just the negative. What happens with the negative, the antibodies find the result band and produce a blue color. Of course, there is always some amount of blue latex that goes through and reaches the test valid band, so with a positive or negative, you still get the blue color in the test valid band. When you get a positive test, you can ship it off to a lab. We did a lot of studies with this in the early days, in 1995, when we launched this product. These were with controls. The controls are stable at room temperature in this cup up to about a month. At four degrees, it goes out to three to six months that the controls can be stable in the cup.

This is to answer your question on precision. Absolutely forget about 25% above and below the cutoff. You won't get 95% with this technology. I would be hard pressed to see any colored technology visual read to give that. Basically, where our lines are here, where we can feel confident, is at 1.5 times the cutoff and half the cutoff. If you want to see 100% of the test showing negative, you have to go down to .25 times the cutoff.

With respect to clinical performance, I could go on for hours with clinical performance. I would just say this, and I will give this to the Committee, I'll refer you to several papers and presentations that were made, one by my group, Jim Towt and company, in JAT in 1995. Amanda Jenkins took a look at the test and presented it at SOFT and she has some materials on it. There was also a group from Helsinki and then Denny Crouch and Dave Kuntz. They did some studies with it. We did support that study at Northwest Toxicology and the Center of Human Toxicology. Bob Willette's and Leo Kadehjian's report should be coming out, too. You can look at their data.

As far as adulterations are concerned, we put our tests through an adulteration panel. These are the adulterants that we have used. For the most part, adulterations do not affect the assay. I will tell you what adulterants affect what assay. For cocaine, if you use 10% Drano or 20% vinegar, you will get a false positive result. For THC, you will get a false negative using 20% vinegar. For morphine, you will get a false positive result using 10% Drano. Amphetamine, you will get a false negative using 20% vinegar. With PCP, you will get a false negative using ammonia or acetone. Acetone is not on this list but we did try acetone in the assay.

We brought some AdultaCheck. If clients want to test for adulteration, we have been recommending AdultaCheck. They can purchase it through us as well.

Manufacturing. These reagents are manufactured just like all the other Roche reagents, manufactured at Roche. We send the reagents out to Ansys Diagnostics in Southern California and they put together the device and they ship it back to us. All 510(k)'s and all the responsibility falls on Roche in terms of the reagents and the product.

Training. Several different aspects of training. When somebody becomes a customer, they are immediately sent a video which describes how the test is performed and what to expect. They are also sent a tip sheet. In 1988, we sent the OnTrack product out for evaluations. People said this is such a simple test that we are just going to send it to 14 sites and collect the results and they should look very good. The research results showed better than 99% correlation with immunoassays. We sent it out to the field and the reports were coming back that it was no better than 50% correlation. In other words, if you flipped the coin, you could get as good of a result. From that day henceforth, we put a lot of effort into training. No matter how simple a test is, you need some type of training. It is best even to make contact with the customers, just to make sure they are running the tests.

What happens, we send this video and we send a tip sheet. That goes to all customers. About 80% of the customers are also trained by a trainer in-house, a trainer that goes to the site. Again, the video is brought, the tip sheet is brought, and also a resource and training guide is brought with them, and the person is given this. It contains history of drug testing. It contains the package insert, and actually, it also contains a quiz. It's a true/false quiz. At the end of the session, the person is tested for it, and if they get an 80% or higher grade, they are given a certificate.

We are taking that one step further. We are putting the quiz on a Web page. All customers will have to fill out the quiz, E-Mail it to Roche. We will correct it. If they get more than an 80% grade, then they will get a certificate that they passed the test.

Don't ask me what we do if they fail the test.

As far as quality control is concerned, we sell a negative control which contains zero urine, and we send a positive control which contains urine at two times the cutoff. There are two levels of recommendation. It is really up to the customer. One recommendation is if you get a shipment in and it's all one lot, you just test it the day you get it. It's room temperature stability. The cup itself is stable for up to two years when it is first manufactured.

In some sites what they do is to test the negative and the positive control each day, so they would have to use two cups.

DR. KIM (Princeton BioMeditech): (Inaudible.)

DR. SALAMONE: That's true. Destroying the drug is something of another issue.

DR. SAMPLE: In an attempt to get all the questions answered that were on the first slides we saw earlier today, what are the analytes in the cutoffs? Any special treatment or handling, FDA clearance, clinical studies that demonstrate product efficacy, recommendations for device evaluation prior to use, and unique features, external PT and any other recommendations for use of the device in workplace drug testing?

DR. SALAMONE: Those are a lot of questions.

DR. SAMPLE: The analytes in the cutoffs.

DR. SALAMONE: Of the HHS panel, it's the cutoffs that are used for HHS. Cannabinoids, it's 50; amphetamine, 1,000; opiates, 2,000; cocaine or benzoylecgonine, 300; and PCP, 25.

DR. SAMPLE: Amphetamine, is that amphetamine and methamphetamine?

DR. SALAMONE: No. It's just D-amphetamine.

DR. SAMPLE: Amphetamine only?

DR. SALAMONE: Yes. I'll refer you to those references. The references are full of clinical data and also I'll supply the Committee with package inserts, which really give a detailed evaluation. It's FDA approved.

DR. SAMPLE: There is no special handling?

DR. SALAMONE: There is no special handling. Just making sure you have the lid of the cup on properly.

DR. SAMPLE: The clinical studies --

DR. CAPLAN: He's going to send them in.

DR. SAMPLE: Any other unique features? What is the recommendation for evaluation before use or is that really covered in running the negative control, et cetera, on each shipment basis?

DR. SALAMONE: Each shipment basis or some labs do it every day.

DR. SAMPLE: Right, but your recommendation is each shipment or is your recommendation each day of use?

DR. SALAMONE: It depends on the site. It's recommended to certain sites to use it -- it's a matter of economy. Ideally, you want to do it every day that you are using the assay, but people don't like using two cups. What I can tell you is that the standard curve of stability of the reagents is good.

DR. SAMPLE: One last question. You touched on it briefly. With the TesTcup, the donor is actually taking the test device into the privacy, if you will, of the restroom or the collection facility.

DR. SALAMONE: The reaction area of the TesTcup is protected by a diaphragm. If someone wanted to adulterate the reagents, they would need to break the diaphragm off and this would be noticed by the tester.

PRESENTATION BY PRINCETON BIOMEDITECH

MR. PORTER: I'm actually with a company called LifeSign, and we are an affiliate of Princeton BioMeditech. I'm actually representing both companies with this presentation. LifeSign is wholly owned by Princeton BioMeditech, and we are really their sales and marketing and distribution arm.

We sell two products or two branded products. One is called Status and the other one is called AccuSign. AccuSign has been on the market for many, many years, sold in criminal justice, workplace, and even in the medical facilities.

Status, although these two products are the same, we started selling the Status product because prior to LifeSign coming into existence, we were actually called Orion Diagnostica. Orion was in business for about 22 years in this country. When we first started doing business with Orion, we let them carry their brand, so now it has come back to us again as LifeSign, so we carry the two products now.

I've got hopefully maybe not a complicated chart here to show how the tests actually work. It's similar to what Roche presented. We have the sample, which is urine in this case, being applied to one end of the device or one end of the membrane. Then there is a pad which contains antibody die conjugate. Of course, this antibody is to a specific drug. Let's say if it's cocaine, it would be specific to cocaine. About halfway down, we have the drug. In this case, it would be cocaine, affixed to it, attached to it, a protein conjugate, like BSA. Then two-thirds to three- quarters down the device, we have anti-mouse, and this is what we would call the control area, control band. There is an immunological reaction taking place, and I'll explain that when we get down into here. Of course, we have an absorbent pad taking off any excess urine.

In a negative sample, the patient would not have any of the drug, cocaine in this case. Therefore, it would not bind with the antibody here. The antibody would then be free to bind with the conjugate, cocaine in this case on the membrane, and therefore what happens is you would get a color band there because it's free. Because this is mouse and the antibody is mouse, you will get a reaction here at the control line to show you that the process took place, there was an immunological reaction taking place here, as well as the liquid reached this point.

Down here is a positive sample. In this case, the cocaine reacts with the antibody. The antibody is not free to bind with this so you get no reaction here and you get no line, you get a negative reaction. Of course, still there is antibody free to bind to the control area and the bound patient sample goes off.

We sell three different devices now. The first device, we call it Cassette. It requires the adding of sample to the sample well area, three drops of urine in this case. We sell these products anywhere from a single drug to ten drugs. In case of the ten, we can only do five drugs on a strip. We need two strips in order to do ten drugs. We also manufacture and sell a stick product where you don't have to do a drop test or drop sample. Simply dip this into the urine. You need to take it out of the urine and lay it flat on the table and then the reaction takes place. We also have another new product here called Status Cup. In this case, the strips are on the side of the cup. This cup sample is put in here. The test doesn't start, however, until the lid is completely closed. We instruct the people who void in these urine cups not to close the lid because the test will start then. Once the test has started, it actually pushes the sample into the area where the strips are, and the reaction takes place on the side of the cup. Of course, that should be done by the administrator of the test and not by the patient.

Obviously, these tests are very simple. There is no special handling of the urine, simply add, in our case, for the Cassette product, three drops of urine, and then you read them in five to ten minutes. Of course, here you are looking for no lines being positive in the test area and a line being present means its negative.

With a stick test, it's the same thing. You dip it in there. We recommend that you hold it between five and ten seconds in the urine, allowing it to be absorbed into the membrane. Again, you lay it flat on the table or counter and in five to ten minutes, you read the results.

With the cup, there is no handling of the urine specifically. You don't have to open the cup or anything like that. You simply close the cap tightly and you read the results, again in five to ten minutes.

These are the drugs that we can detect with our devices. Methamphetamine. Our cutoff is 1,000. Opiates, of course, we have two devices we sell now, one at the old cutoff value of 300 and now the new SAMHSA guidelines at 2,000. Actually we are selling products at both of these cutoffs. Cocaine, 300. Marijuana, cannabis, 50. PCP is 25. Amphetamine down here is also 1,000. These are basically the SAMHSA guidelines for recommended cutoff levels.

We have devices for benzol, for barbiturates, for methadone and tricyclic antidepressants. I guess you might say this is kind of the industry standard for the cutoff on these particular drugs.

This is part of our package inserts for our products. In this particular study, we compared the status of the AccuSign product to GC/MS. These are the levels, the range of drug levels that we had that were GC/MS positive and you can see in this study, it was 100% agreement.

We also compared the products to other rapid or screening tests. The EMIT was an instrumented system, the Roche OnLine product and the Biosite Triage assay.

In this case, we had really good agreement except for maybe amphetamine and methamphetamine, which was 87%. We do know some of these tests, like the Triage and the EMIT, they don't differentiate between amphetamine and methamphetamine, and that may be the reason why we didn't see quite as good agreement with these particular procedures.

All of our products are FDA approved in all the formats. Again, we sell them in singles and each individual is FDA approved as well, combinations, multiple drugs, the Status Cup as well as the Status Sticks. The cup and the stick are only cleared and we only market them for six different drugs. Whereas the Status DS, as we call it, they are cleared and marketed for up to ten different drugs. We do not market any products that are not FDA cleared.

Unique features of the product, as everybody has said before, no special treatment. Just add the sample, depending on which test you like, we have designed these products to meet the needs of the customer, whether it's a stick or a cassette or a cup. It's basically add the sample and it's a walk away test, and then you come back and read it in five to ten minutes. It's rapid. It's visual reading. It's non-instrumented, at least at the present time, although we are looking at instrumentation to take the subjectivity out of this at the present time.

In essence, it really doesn't require highly skilled professionally trained people to run these tests, as has already been attested to in some of the studies. Someone who is trained properly can run these tests, because they are simply looking for no color or color. Our tests are unique in the fact that you can pick and choose what drug you want to test for, anywhere from one to ten.

The kits are stable at room temperature up to about 15 months. Usually when a customer gets them, they are about 12 months. You can ship them and store them at room temperature.

They are highly sensitive and accurate. You have seen reports today, compared to the competitive products, they look pretty good.

We did some studies on saying how the product looked from zero to the cutoff level. We did some inverse linear studies because what happens basically is the color decreases as the test approaches the cutoff. These studies show that as the sample gets closer to the cutoff, the color line slowly disappears. It's pretty linear. This is log logigraph, of course, but it's linear regression.

What are our recommendations for QC? Actually, CLIA guidelines. We are a GM facility and are regulated by the FDA, as well as Princeton BioMeditech is a GM facility. We try to follow HCFA and basically say for these types of devices on a daily basis, you can use the procedure control as your daily QC. We recommend certainly you take a look at that control line to make sure it's there. The clearing in the background, these are actually CLIA guidelines. The clearing in the background behind that control is suggested, the negative procedure control.

In addition to that, we like the other manufacturers do recommend external positive and negative controls. We say we recommend them according to good laboratory practice. If asked, we would recommend similarly like the other manufacturers, run them with new lots and with new shipments. If you get the same lot in the second shipment, you should test it again.

As far as adulterants, we recommend obviously the visual inspection, like everybody else does. Our tests, including the cup, has a temperature strip on it. You would want to take a look at that temperature strip to make sure that complies with the guidelines, which I think is 90 to 100. Adulteration test strips, we like the other manufacturers do sell a Delta Check. Delta Check doesn't test for specific gravity, but we recognize that people are using all kinds of adulterants. Sometimes you can see it because you will see some precipitant on the membranes. We usually get those in because people will say something happened to the test, but in reality, somebody put something in the urine.

Some of these things, if they do get really extreme, like pH that is really low, we have done some studies recently to show that the levels -- we don't even get a control line if it gets around 2.0 or something like that. We are trying to work more on these adulterations to see what the effect of various adulterants is our device. For the most part, they don't affect the control line, which is unfortunate.

As far as recommended training, what we provide, we certainly recommend training on these tests, although they are very simple. People do need to see them and need to run them before they actually start running patients.

We have a training video. We also encourage on-site training of our representatives where possible. We pretty well cover that as best we can because that pretty much eliminates a lot of the complaints you might get from new customers, and we also have an 800 line, like I'm sure the other manufacturers do, to answer any questions when somebody is running the test. The 800 line gets used a lot for first time users.

DR. WALSH: Obviously, this product is also distributed under the name of the Syva Rapid Test. If you could tell us what other names the product --

MR. PORTER: Just the three names.

DR. WALSH: You showed us the Status Cup and the Status Stick. Are they also marketed as AccuSign Stick and AccuSign Cup?

MR. PORTER: No. Right now, we are marketing them totally as Status products.

DR. WALSH: I asked Ken McNeil if his product was technically the same. Mr. Fortner implied that maybe some of the manufacturers are getting rid of the product that don't meet the specs of their competitors.

MR. PORTER: We don't do that. We basically only sell the product to Syva under an OEM agreement. They are very good in testing and making sure that the product we sell to them is identical to the product that we sell under the AccuSign and Status brand. No, we don't do that and we don't ship it to Europe either or Southeast Asia.

DR. SAMPLE: You said you knew what one of the questions is. Do you know what the answer is?

MR. PORTER: Is this going to be the 50%, 25% above cutoff? We strive very hard to meet the guidelines, try to do 25. I can say that probably like everybody else, realistically, 50% above and 50% below the cutoff is more realistic than trying to do 25 above and below. We do provide controls to our users as well. Interesting enough, we have negative and we have one that's between 25 and 50% above the cutoff. We test that with each lot that we manufacture and it works fine. In addition to that, we have another control that is at two times the cutoff.

DR. SAMPLE: I would agree. In my experience, I've never seen a plus 25% control fail to work on anybody's device. It's the minus 25% controls where I've tended to see more difficulty. My second question relates to the Status Cup. Again, we have a situation where the donor can take the test device into the privacy of the restroom. What precautions, if any, do you take to ensure that no tampering can occur with that test device?

MR. PORTER: The test device itself, I have a cup down there. You can't get to the strips, number one. We tell them just to give it a couple of turns. It looks like this. There is a chamber here where a certain amount of urine goes into and the rest of it stays in the bulk area. They can't get to the strips. The only thing they can do is turn it all the way down. If they do that, they are going to start the test and then of course, the administrator knows they have started the test. You can read them within five or ten minutes. In excess, because they don't know the time, this probably wouldn't be any good for them. They would have to collect another sample because they wouldn't know how long it has been sitting in there. No, they can't get to the actual test.

DR. SAMPLE: Unless they lift the plunger up.

MR. PORTER: Try to lift the plunger up.

PRESENTATION BY MEDTOX LABORATORIES

DR. McCOY: I titled this on-site drug testing meets the existing laboratory based model. I know most of you, the DTAB group, were not expecting coming to think that you would approve a system today, but I think you will be changing your mind after this presentation. I say that only a little bit tongue in cheek. I do bring a little bit of unique perspective to this being in the laboratory industry for a very long time and now the last three years or so also being a device manufacturer. Some of that I think will come through the presentation today. I did try to make it easy as some of the other presentations have, to go through the 12 questions that we had and give specific answers to each in turn. Yale and then Barry gave us some new questions today and I will try to address those within the text of my talk.

First of all, ours is also an immunoassay device and it is patented. It's a single step, lateral flow immuno-chromatographic device. It is essentially identical to every other drug device manufacturer's device. I will just point out maybe a little more emphasis to the line's lateral flow here, because there have been some questions about if the same strip in a different device is an identical device, and I think most of us in the industry would say that the flow characteristics are very dependent upon the holder itself, so the strip is part of it but the pressure points and so on of the holder can indeed make a big effect, and that is an issue that should be considered.

What are the performance characteristics of the device, referring to cutoffs and so on. They are essentially targeted at the SAMHSA cutoffs. We will say the opiate 2,000 threshold for us has been submitted to the FDA but it is still pending approval. I feel fairly competent saying that FDA approval will come before the regulators approve it for mandated drug testing.

The special specimen treatment, two drops of urine. Profile II is 510(k) cleared, and also we have a family of other products that are also FDA cleared.

I want to mention the early morning presentations were outstanding in talking in general about the field and the performance. Bob Willette's studies referred to some of our early generation devices, EZ Screen, Profile and Verdict. Profile III is now our third generation of device. We have been in this field for quite a while.

Provide the clinical studies that demonstrate product efficacy. First of all, we do believe -- we know the SAMHSA and other certifications require FDA approval for the immunoassay reagents used in certified laboratories. We feel that standard is a good one and should continue to apply. Also, it is worthwhile knowing there has been some questions and answers earlier about variability and so on. I think it's important to point out that FDA does enforce compliance of specific thresholds and so on. We do feel very strongly about that. Also, a lot of the clinical study data is available in the package insert, which will be given to each member of the Board, the working group.

A few other comments about efficacy and clinical studies. I think many people in this room are familiar with the fact that the laboratory based liquid immuno reagents are not identical to one other. I don't think we should assume that the devices used for on-site testing will be or perhaps even should be identical either. They do have some different target analytes. Dr. Sample has asked a number of questions in the amphetamine area. We will talk more about that.

We know they have variability in the cross reactivity of the different metabolites within the class and then also different analogs within each pharmacological class. Those of us who grew out of the clinical laboratory area are very familiar with Dr. Westgard and his rules of quality assurance and so on. I want to take comments that he made recently in the Journal of Clinical Chemistry, talking about the main purpose of medical comparisons, to look at the decision points, what are the important decision points, and are we testing with our method of comparisons at those decision points. I would argue that similar to what Dr. Lappe said this morning, these point of care devices are being used to identify negative specimens in the field. Are our method comparisons evaluating that or are they evaluating something different.

To evaluate the efficacy in the field, we think the evaluations and comparisons should be made with real specimens. That means fresh, real, with all the naturally occurring metabolites, being able to correctly identify the specimens as negative and positive compared to other existing methods, be it screening methods or be it confirmation methods, not by an artificial method.

I know that Dr. Willette and others have put a lot of work in the different study designs in measuring devices, and a lot of that have been study designs based on the way to analyze quantitative laboratory tests. I'm not sure that really is the best way to do these evaluations. I might point out this essentially is the way that the studies for FDA submission are performed and that is the information contained in the package inserts.

I'm going to talk a little bit about our own experience with the devices, talking about clinical studies and product efficacy. We have marketed now several thousand Profile II devices with automatic laboratory confirmation. I will have more to say about that.

In our laboratory based testing, 6% of our specimens screened positive at the immunoassay level and go onto confirmation. We do have more specimens going out to confirmation from the field than we do from the laboratory. I think there are a couple of reasons for that. One is we are recommending individuals to send specimens into laboratory whenever they want to. In training periods, they send high levels in. Many times they send specimens in even though they clearly test negative in the field for confirmation just to evaluate that everything is working properly, especially during the training intervals, and over time, that number might go down a little bit.

There was a question this morning from an individual on the floor as to what is the place for these point of care devices. I think the answer is involved in that question as well. There is now a new population of employees or applicants that are being tested, fast food restaurants and hotels, construction sites, temporary workers and so on. It's a new group of individuals being tested for the first time that may indeed be testing positive at higher rates than those that have been tested traditionally in the laboratory. I think there are more than two reasons for that difference. Those specimens that do come into the laboratory for confirmation, 67.9% of those specimens are ultimately confirmed by GC/MS. That may not seem that high, but compared to our laboratory base, where 73.6 of those that are screened positive by the laboratory are being confirmed. If you compare now 67.9% to 73.6%, that's 92% agreement. Again, I think a lot of the specimens that are being referred for testing are not truly being referred to testing because they screened non-negative in the field. Similar again to a laboratory model, when THC and cocaine screen positive, they are more likely to be confirmed. In areas where we have amphetamines, opiates, phencyclidine, where now we have analogs involved and other reactivities, it's less frequently confirmed.

Overall, when Profile II is used in the field, referred to the laboratory for confirmation, 72% of all specimens screened are confirming positive. Compare that to our or any other laboratory's confirmations, I think some things become evident. That's a good confirmation rate. Therefore, we feel that in real life, false positive is not a problem because we are ruling those out with GC/MS. False negatives in the field are not a problem because we are seeing a strong percentage of positives, and also we believe that with that strong positive rate, that we are also not missing an undetected adulteration.

That leads directly into question number six. What do we think about adulteration testing? We have very strongly believed in dilution testing, adulteration testing for a very, very long time. We think that holds for point of care testing as well. Therefore, we believe that adulteration testing should be done at the collection site whenever a point of care device is used.

How should the device be evaluated prior to use? We believe that somebody should only use an FDA approved device. We believe that the integrity of the device should be inspected by looking at the integrity of the foil wrapping of the device, the expiration date should be verified and the lot number itself should be recorded.

What are the unique features of the product? Here I'm going to begin talking about more than the device itself. The devices do detect the presence or the absence of drugs. We really put it into a system that works for workplace testing. We call it the Profile II defensible system, and what I'm going to be talking about could be duplicated by anybody. It doesn't need to be unique to us. Absolutely anybody could put together this program.

Rather than just selling the device, we sell a system that includes everything that needs to be done for proper collection, the wrapped collection cup, the temperature detecting device, split specimen containers, the chain of custody, very important, the tamper evidence seals. It also includes automatic confirmation of every non- negative and it also includes MRO review of every specimen.

I mentioned electronic reporting to the medical review officer, every negative specimen in the field or any positive specimen that comes from the laboratory. Either way, it goes directly to the medical review officer, just like laboratory based. The negative results come directly from the collection site to the MRO. This comes through our own electronic reporting system. Obviously, the non-negative specimen would be forwarded to the lab directly and then reported out. The advantages of this would be both the MRO review is still involved but also we have all the data in one place for the statistical management for employer use, et cetera.

What are our quality control recommendations for the device? Similar again to the discussion we have been hearing. We do have a control line on each device. This does evaluate the immuno reactivity. It does do some validation to the recent storage conditions and again, it demonstrates an adequate specimen volume was applied to the device itself. We also believe in suitable negative and positive external controls. In this regard, since we are opening with a new comparison chart for all the methods of analysis, we think that the existing quality control standards for laboratory based is too intensive. Currently, we are required to run 10% of all specimens. I think our experience has been it need not be that high, particularly the way it is being performed today. On the other hand, we do believe in quality control and at point of care, perhaps 4%, which would be one device in 25. Definitely without exception and when there is a lot change, we believe there should be quality control documentation, and then perhaps something like once a week or some other parameter based on the testing volume would be appropriate for quality control.

We do run on-site testing ourselves. Since we are a laboratory, we do run negative and positive controls each day that we run the device. We do have the two level control on a daily basis and we have had absolutely no difficulty meeting any QC standards doing that.

What are the external performance testing recommendations? We have always encouraged the periodic sending of a negative specimen to the laboratory. From time to time, we think that's a very good idea. We ourselves again at our own site have participated in the CAP program for urine testing for over a year now, and as I recall, those results are listed as satisfactory or unsatisfactory. We have been consistently satisfactory in each of the rotations.

I put RTI in here. It could be RTI or Duo Research or any other number of vendors. We also think there may be room for an outside agency to have specific proficiency testing programs for this area. Perhaps that could be coordinated through the manufacturers themselves.

Question 11, what are the recommendations for training on the use of the device? We have for some time offered Web based training. We are aware, however, that not everybody has access to the Internet, so we also offer these materials in hard copy. For quite a while, we have had certification exams available through the Web and again through hard copy. If people are doing it through our Web site and they have a printer hooked up, they successfully complete the exam, they can print out their training certificate. Otherwise, we will be happy to mail them the certificate following completion of the certification examination.

Question 12. How do you recommend using the device in workplace drug testing situations? Controls, reading, automation. Again, we do believe in mimicking the very well working laboratory based model, where you have well identified donors, you have well identified tamper evidence specimens, you have correctly completed chains of custody, and in this regard, with our electronic reporting, we review the images of the chains of custody coming in, documenting electronically that all the required fields are filled out so that before releasing any results, negative or positive, we document that we have a proper chain of custody.

We believe you need the appropriate precautions against dilution or adulteration. We believe in the automatic confirmation of all non-negatives without any choice, and we believe in MRO review.

MEDTOX is currently working on a reader. Some comments about that. We believe that the strip, the holder and the reader itself does need to be FDA approved prior to any use or sale. We are not aware there are any readers that are approved for this application. We do feel there might be a convenience, the reader might be a convenience at some collection sites. On the other hand, we do have a precedent of not needing a reader for visual tests, breath and saliva. Alcohol screening devices, some of which require a visual examination, those devices do not require a reader to be approved devices. We don't see any reason why drug screening devices should be handled any differently than that. Therefore, we feel that a reader is not necessary for valid drug testing screening. Also related to this comment is that we know that any positive result will be screened and confirmed by a certified laboratory.

One other question that wasn't directly asked but is a comment that we hear, is there a lack of confidentiality with doing on-site drug testing. We again think we have a precedent set with alcohol testing. One BAT performs the test, records the result. Of course, it is repeated if it's not negative. On-site drug testing we feel is safer than that because at no time will a positive result be released by a collection site, which is different from alcohol testing.

I was at one of the early DTAB meetings when new nomenclature was added to the toxicology world, the Bubba factor. Talking about the Bubba factor, we recommend that a drug test, even though it is performed on-site, not be performed directly in front of the donor. We think that it may be useful in many cases to do adulteration tests either in front of the donor or where the donor is nearby, in case a second collection wants to be performed almost immediately. We do believe it is best to excuse the donor, at least to another room if not outside the facility, before completing the test. Again, we have chain of custody forms, specimen sealed in front of the donor, and then tested immediately thereafter.

MR. LINEWEBER: Would you clarify the statement you made earlier in your presentation about not being concerned about the adulteration when it goes to the laboratory? Would not the nitrite mask the THC?

DR. McCOY: Correct. I probably was not completely clear. There are two statements that I should re-state about adulteration. Number one, definitely adulteration testing should be performed at the point of collection, when the test is performed at the point of collection, without question. My second remark, with doing that screening immediately and then our final results, we believe we are not having a problem with undetected adulteration because of the high positivity rate and the fact that we are doing the screening. The other interesting thing that we found and Dr. Collins is in the audience, she can either confirm or deny that, we have also found that we have less frequent attempts at adulteration in the field than we do in the laboratory. This again may be due to the fact that we are mainly talking about a new population that doesn't have the history and training of performing drug tests, but we have less frequent attempts of adulteration in our field tests.

MR. LINEWEBER: In the railroad industry, we have a lot of adulteration, it appears to be adulteration, going on in the field right now. The proliferation of that, that is our concern.

DR. McCOY: We also have experience defending litigation in the railroad industry with adulterated specimens, but again you are talking about people who have been trained in giving drug specimens for a long period of time.

MR. SMITH (Chimera Research): As far as doing adulteration testing, if you are not doing it right at the site and the specimen is collected and sent to a laboratory and this person has left the building, and of course, they are not around any more days later when you are doing the drug testing, isn't it sort of a moot point? If you do adulteration right there at the site, that gives you the chance to hold the individual right there, maybe it's a low creatinine, you can hold them on site. If you are waiting a few hours or a few days, you are not going to be able to do that.

DR. McCOY: Perhaps I was again misunderstood. I am recommending immediate adulteration testing.

DR. SAMPLE: At the beginning of your presentation, you alluded to Yale's two additional question and my additional question and also in your presentation, you said you would address the issue of amphetamine and methamphetamine. I was wondering if you could answer all of those at this point.

DR. McCOY: MEDTOX does the research, develops the patents, does the manufacturing, does from A to Z for development and manufacturing of each of our devices. We do not distribute anybody's device other than breath alcohol tubes we don't manufacture ourselves. The amphetamine, our device is specific to amphetamine itself. We feel that's appropriate and we feel that the clinical studies using real specimens bear that out. I think with the challenge of specimens that meet SAMHSA guidelines for containing 200 nanograms of amphetamine and methamphetamine, that you will see consistent positives with our device.

DR. SAMPLE: And on the plus and minus 25% or plus and minus 50%, where is your device currently and where you do think you can go?

DR. McCOY: If you look at Dr. Willette's early studies, I think you will see that our early two generations of devices performed rather well compared to a few false positives and false negatives. We now have a newer generation. Part of my discussion of talking about it, I think we need to consider a different model of comparing methods. I don't believe plus or minus 25% is an appropriate model for this industry.

DR. SAMPLE: I understand that, but could you still state where your devices currently perform?

DR. McCOY: We are pretty good within 50%; 25% is at the edge.

DR. SALAMONE: Just a comment on adulteration. This doesn't necessarily go for instrument based assays where if somebody puts in an adulterant and it's not tested until 12 hours later, but with these on-site devices, when we have done the adulteration test, if you spike in the adulterant within an hour's time of testing, it really doesn't eat up the drug. I was just thinking about it. With the nitrites, for instance, none of the drugs were eaten up in that short of a time. I think there is some time line. I think adulteration with temperature, if it's stored at a higher temperature, the adulterations will eat up the drug more often. I don't know if you have seen the same thing but we just don't see adulterations eating up the drug.

DR. McCOY: I wonder, Sal, if you are thinking about half of the problem and maybe that is because you are a device manufacturer. There are two questions. One, will the adulterant, the addition of an adulterant, affect the immediate screen test, whether it's done on the on-site environment or in the laboratory, and then the second, will it ultimately confirm positive in the laboratory, and then still further, will it confirm upon re-test challenge. I think it's very likely that for many of the adulterants, there won't be time for any deterioration of the drug molecule itself for the on-site test, and perhaps not even for the initial screening assay. I think some of the most recent guidance regarding re-test analysis in adulterants are quite good in that regard. One problem we do have in a laboratory is by the time specimens have sat around for a few days while waiting for a request for re-test, now there is plenty of time for deterioration of the molecule and the presence of pyridium and chlorochromate or nitrite, or something to affect the results. We definitely need to detect them somehow.

PRESENTATION BY CHIMERA RESEARCH

MR. SMITH: I just learned that when I got here today that our materials didn't make it, but I think we will work our way through this anyway. Chimera Research started doing adulteration testing in 1989. We literally founded the industry, got it started, have been doing it ever since. We actually publish and put out booklets and training materials on what adulteration is all about.

There was a real good question. MEDTOX does an excellent job screening for adulteration testing. These guys do everything in the laboratory and I believe at the collection site that you probably could do for it. My question was this to him, and I think it was misunderstood. He said you would do point of care adulteration testing if you are doing point of care drug testing. I think you should do point of care adulteration testing even if you are not because the individual still gets away from you if you are waiting to do the adulteration test. That was the point.

This is the AdultaCheck. It's a dry chemistry test strip. I'm going to show it to you. It's very simple to use. Dip sticks have been used in the industry for years, so everybody knows how to use them. There is no real secret to this. You just dip it in the urine. That's basically how it is used. It's real easy to use. We would dip it in urine, compare it to the color chart. The thing about using something like this with lateral flow, you get a yes/no answer typically, but with dip sticks, you have a band of color, so not only do you get the cutoffs that you would normally have, the SAMHSA cutoffs, which are used by the AdultaCheck, but there are other colors.

We actually at Chimera believe, for instance, that the SAMHSA guideline of 3 and 11 is the pH cutoff is a little wide. You are probably missing about 50% of the pH's that would probably adversely affect adulteration, I mean it would adversely affect the drug test. They should probably tighten down that guideline.

The other thing was mentioned earlier, specific gravity. Why don't we have specific gravity on the dip stick? There is about four or five published articles, peer review articles, for specific gravity, that have shown if you have a specific gravity of let's say less than 6.0, it will cause the color block to move up one, or if you have a specific gravity of greater than 7.0, it will cause the color block to move down one. A normal urine will adulterate the test. Why would you have an adulteration test for specific gravity that in itself could be adulterated by a normal urine. That's why we didn't do it. I think I've heard of some being out in the industry but it doesn't make a lot of sense.

The testing principle for the AdultaCheck, the assays are all read immediately. The performance characteristics, as I mentioned. The cutoffs are all there for the SAMHSA guidelines and the SAMHSA cutoffs. For pH, it's 3 and 11. I don't have it right in front of me. I believe the SAMHSA cutoff is 3 and 11 for pH. Is that right? We have the nitrite. We have the SAMHSA cutoff for nitrite. We are also a little bit lower.

There was a really good published article that talked about how strong of a nitrite could a person produce in a normal urine, that it was something that was actually added. The highest they could ever get in urine was to about 130 milligrams per deciliter. The SAMHSA cutoff is 500, which is way up there. There is no way this is a normal sample. This is somebody obviously adding nitrite. We have a 150 cutoff on here but we also have the SAMHSA cutoff for the nitrite. Then there is glutaraldehyde according to SAMHSA and according to us. Any glutaraldehyde in urine is a tampered urine. Humans do not urinate glutaraldehyde. There is no reason to have any in there. Of course, there is the pH.

That is basically the performance. The specimen treatment, there is none. You would just use a random urine sample, that would be perfectly adequate. You would either dip it into an urine or you could dropper it onto the urine. They are all read immediately. The product is FDA cleared.

There are clinical studies. There is a published peer review article on AdultaCheck. That was published in JAT. The article highly recommended the use of AdultaCheck. It's been pretty popular. It's really the only product out there that FDA approved for this. There will be some more things that we are coming along with, so there will be new things. There is all kind of stuff coming up in adulteration testing, new adulterants, new problems, new issues.

We recommend that you do test for adulterants. That's our recommendation. How is it evaluated prior to use? We would recommend that you evaluate the AdultaCheck just like you would any other new product you are bringing into your facility, your laboratory. You may run 20 replicates, do it within day to day quick studies to see how it is working for you and if it meets your in-house quality control guidelines, then it's ready to go. We haven't had anybody reject it on that basis. It should work fine for you.

There are some unique features. With the AdultaCheck, you are able to do a series of adulteration tests. You can do it quickly. You can do it on-site, minimum training. This is a dip stick. It has been in the industry for 30 to 40 years. People have been using dip sticks forever. It is not going to take anybody a long time to understand how to use this assay.

Training on use of the device. That's basically it. Recommendations for use of the device in a workplace. We would recommend daily QC. We do sell a control set that goes with it. It's a positive and negative control. It would last probably 30 to 60 days. It like the AdultaCheck is stable for about two years at room temperature. They are very stable for a real long time.

We do manufacture the AdultaCheck. We manufacture everything that we make. Chimera has about 80 FDA approved products. Everything we make, we manufacture in-house. Of course, a lot of our product lines are in the liquid adulteration line for automated analyzers, so we sell that to a lot of the clinical labs.

Do we distribute to other manufacturers? We do OEM the product. We do distribute it to a lot of people. It's worldwide distributed.

MR. WILLIAMS: You lost me on your discussion of gravity, specific gravity. 1.00?

MR. SMITH: Here is what happens. On a specific gravity strip, for those of you guys that can recall it if you have ever used it a lot, specific gravity on a strip is typically 1.000, 1.005, 1.010, it's a qualitative sort of cutoff that they will have for specific gravity. These studies were done a few years ago. There is probably about five or ten studies. They were all focused right on specific gravity, dry chemistry measurement of specific gravity on a dip stick. What was happening was if you had a urine that was pH like 5.5 or 5.0, it will read falsely high. Therefore, you could have a creatinine value of zero but an acidic urine of a pH 5 and a specific gravity is going to read like 15, 10 or 15. You wouldn't catch them diluting the sample. The pH is adulterating the value of the specific gravity on the strip. It's not a good method to use for dry chemistry, for adulteration. For regular clinical samples, it's fine.

DR. KADEHJIAN: This question actually has to do with all devices that are going to be used to make an assessment of an urine specimen. As toxicologists, we are all told that you should never ever put anything into a specimen, less you risk contaminating it. One issue that's come up with dip stick type tests, whether they be adulteration tests or drug tests, is that this should be done on an aliquot poured off. For you and all the other manufacturers, have you done studies to demonstrate that you in fact can put that into the primary specimen without any concerns of specimen contamination. I bring that up because there was a recent paper in Clinical Chemistry that addressed falsely elevated results and they noted the history of that occurring. I personally believe that it's a non-issue, but I'm just asking about what your studies are on your device.

MR. SMITH: Unfortunately in drug testing and in adulteration testing, it gets a lot of legal attention lately. The courts are upholding that if people are tampering with urine, it's almost worse than having a positive drug test. It's like tampering with evidence. Getting to what you are asking about, there is actually a lot of published work on the old DeMile strip and the Manheim strip for urinalysis, do they do anything to the urines and actually they don't. They have never really ever affected anything along clinical lines, so the assumptions are made that they are not. We just basically did studies into what had already been done.

DR. CAPLAN: That included drugs?

MR. SMITH: Yes, that included drugs. There is not enough of it coming off to affect anything.

DR. LAPPE: I think it's really the same as Leo's question, is the recommendation to dip the adulteration check into the forensic sample?

MR. SMITH: For people who are doing forensics, like I was saying earlier, you can dropper it onto the strip with a dropper, just like the same dropper that you are using to put on your lateral flow device. You would wipe it off anyway. The way that the instructions read, you dip and then you run the edge along the cup so you can get rid of the urine or you can use a dropper, put it on there and do the same thing. You can dropper it on there without ever putting the strip anywhere near the cup, using a dropper.

DR. LAPPE: Is it contraindicated to dip it into the sample?

MR. SMITH: No. We feel that it doesn't affect it one way or the other. Forensically, yes. Attorneys are not wanting you -- I'm sure you guys know about this. You are not supposed to put anything in the specimen. In that case, if it's for that type of testing, I would say dropper it onto the strip. It works fine.

DR. SALAMONE: Does your agglutaldehyde test react with any aldehyde?

MR. SMITH: It reacts with other aldehydes, but it's very specific for agglutaldehyde. It's basically very sensitive to aldehyde. We have done extensive testing with other types of urinary ingredients and stuff and we are not getting any interference that will tick over the color markers that we have for the presence of aldehydes. When that aldehyde hits, it's very low concentrations, very sensitive to very low concentrations.

DR. SMITH (SAMI): I have a couple of questions in regard to the application of the AdultaCheck specifically to point of collection testing. Number one, is it your position that the AdultaCheck, if it in fact shows a positive finding or that adulteration may have occurred, that the specimen be considered invalid and a re-collection take place, or is it your position that the specimen should be sent onto the laboratory where other confirmation, I use that term loosely, but other testing for the specific adulteration issue should be carried out? Which is your recommendation?

MR. SMITH: We actually recommend that you confirm it by another method if possible, the adulterant. If you suspect it, we recommend that you send it to a laboratory or in-house, you would check it with a pH meter. If it's agglutaldehyde, you would send it in and have an agglutaldehyde assay run on it, a liquid agglutaldehyde assay, or you could run it by GC/MS or some other type of methodology. We do strongly recommend that you confirm it by another method.

DR. SMITH: The reason that I ask some of these questions is that now with adulteration or substitution of an urine specimen becoming as I think you alluded to, potentially a specific legal finding with regard to refusal to test, regulated testing or in other circumstances, it would seem to me then that this product would have to be subject also to some of the same kinds of standards with regard to how accurate is it at the 500 cutoff level for nitrites, for example. Is it within plus or minus 50% of that cutoff level, are there false positives. By that I mean circumstances where the AdultaCheck would fail to confirm using liquid or other. I think it becomes an issue, if you are going to use it for some kind of potentially sanctionable issue. If you are going to use it simply to require another collection under direct observation, then I think the standard may be somewhat different. Those are my questions relative to the product.

MR. SMITH: As far as the strip goes in in-house evaluations and how we do those, we follow the NCCLS guidelines, and for dry chemistry test strips, they have a guideline for determining a cutoff, plus or minus one color block. It is a semi-quantitative assay and that's how you determine it. I agree with you actually as far as how you would recommend to do those guidelines. It is a good screen to do adulteration and in a court of law, I've testified as an expert witness on forensic tests, before I started Chimera, I was a forensic toxicologist, and two methods are typically pretty good, two or more, and the dry chemistry will stand up as a method itself, and then whatever other method you decide to use will probably stand up in a court of law, especially if you did two methods.

DR. CAPLAN: Can you comment on what you believe the accuracy of colorimetric type tests standing by itself are compared to other types of tests? In adulteration testing, we have approached a new area of limited resources and analytical capability. We are not doing GC/MS. We are not doing three tests, et cetera. How good do you think the colorimetric tests are in comparison to the validity of drug testing in general or how important is the confirmation, and as a corollary would be alcohol. For alcohol, we know. For alcohol, we accept one method generally. You might do it twice. How do you feel about the clinical chemistry parameters in that regard?

MR. SMITH: The thing about dry chemistry colorimetric assays, they are subjective. The individual reader is actually very good. You can get very good and accurate with it. What if you didn't know that somebody was screening and they didn't know they were slightly color blind to some particular color. There could always be that kind of issue. They actually tend to be very accurate. In the independent study, the peer review published study, it tended to be very comparable actually to the quantitative methods in every case of actually detecting adulteration. It was actually very accurate.

DR. CAPLAN: I didn't mean from the point of adulteration. I meant from the point of individual specificity of a colorimetric assay. Do you feel the ones that are used -- we have been doing creatinine assays for years. Is that considered by you a highly specific assay or what kind of things might interfere with these kinds of assays?

MR. SMITH: For this purpose, I think it's highly accurate. With the Jaffe method for running a creatinine, there is some classical interference problems that have sort of always been around. If you have a urine and it tested to be 10 milligrams per deciliter or lower for creatinine on the dip stick and then you run a Jaffe, which is a different method from this method, and you get a 10 or 8 milligrams per deciliter creatinine, you can pretty much put your money on it. The other thing to remember is with creatinine, for instance, we are looking at a very dilute urine. The chances of interference are less because the concentrations of everything in an urine are less. It's a very dilute urine, so the chances of even worrying about interference shouldn't even be there just because it's a dilute urine. If it's a high specific gravity urine, high creatinine, got a lot of things dissolved in solutions, then I might be a little bit more worried about tampering or something like an issue where the color wouldn't develop like it was supposed to.

DR. LAPPE: This is a general question, just curious. Don't you think that with point of care testing, especially forensic testing, that the adulteration and integrity test has to rise to the level of the drug analyte test because we are making decisions about what to call a negative and what to send onto the laboratory. Should they really be semi-quantitative or qualitative tests or do we need a better threshold or something to measure what is a valid sample?

MR. SMITH: For instance, drug test, lateral flow drug test, you get a positive in a screen, you are going to send it to the lab for confirmation. With an adulteration test, you get a positive in a screen, you are going to send it to the lab. There are now available for just about every adulterant on the market, we market them ourselves, quantitative FDA approved assays that are very specific and get it down to the levels that would be very sensitive. It's sort of the same thing. You are quanting the drug after you have screened, and we can quant the adulterant after we have screened. It's there. The cutoffs, definitely, I think need some work actually.

DR. McCOY: If I could respond just a little bit to some of these questions about screening for adulterants and confirmation. I feel strongly that any presumptive finding of adulteration or dilution does need to be confirmed. I have always felt that way, felt that a forensically defensible test needs to be performed, whether it be agglutaldehyde, nitrites, pyridine, chlorochromate or low creatinine. In the case of creatinine, we don't have fantastic defensible methods developed, and in that case, we have the two different analyses, the one being creatinine, the other being specific gravity that we perform before declaring a specimen to be dilute. For each of the other specimens, I think if it is done on an on-site basis, it definitely needs to go to a laboratory for confirmation. It may indeed, depending on an employer's policy, it may be a great idea to collect a second specimen right then and see if the second specimen would give you additional information. That first specimen should indeed be sent to a laboratory for forensically defensible confirmation tests, no matter what the adulterant might be.

DR. CAPLAN: I'm sure this is an area we are going to be hearing about and hearing a lot more with. PD-37, PD- 35, two specimens, two methods, one quantitative. That has been sort of the de facto criteria. There was latitude put in there for ones for which we have no specific cutoffs, et cetera, agglutaldehyde, but quite frankly, the labs are on their own until we get better information.

I know it's an area of great interest and concern and we probably are not going to figure it out today.

DR. CAPLAN: Are there any other questions that anybody wants to particularly bring up? We will have some time tomorrow hopefully at the end of the morning and probably in the afternoon for any further discussion.

[Whereupon, at 5:25 p.m., the meeting ended, to reconvene the following day at 8:30 a.m.]

AGENDA (October 6)

Review of 1989-1990 On-Site Drug Testing Laboratory Study - Dr. Donna Bush

On-Site Drug Testing Laboratory Experience - Dr. Leo Kadehjian

U.S. Postal Service Pilot Project Update - Ms. Julie Murdoch

Corporate Experience with NIDT - Dr. J. Michael Walsh

Drug Testing Performed at Collection Sites - Ms. Wanda Boone

MRO Experience with On-Site Drug Testing - Mr. Ted Shults

Questions and Answers

Manufacturers and Distributors Presentations

REVIEW OF 1989-1990 ON-SITE DRUG TESTING LABORATORY STUDY

DR. BUSH: I would like to start with a brief review of where we have been with the issue of on-site drug testing in the past. It started a long time ago with a consensus conference, when this Division and this program was still in NIDA and before it migrated over to SAMHSA. In 1993, there was a report commissioned or actually a working group formed where we needed a work product. We needed an evaluation of on-site drug testing in the workplace as it was at that time. The consensus conference was in 1989. There were many issues common to on-site drug testing then that are the same issues occurring today. This report was generated in 1993.

The working group that was evaluating on-site drug testing and the sites at the time included representatives from organized labor, private business, and Federal agencies. The purpose was to obtain a snapshot of some on-site drug testing practices at that time and to provide observations and recommendations to the DTAB and NIDA (now SAMHSA).

This effort was policy neutral with focus on testing practices. Eleven sites were selected and volunteered, and these routinely performed on-site drug testing. They volunteered for site visits, otherwise known as inspections, and a variety of industries performing on- site drug testing at that time, military, labs, industries, and criminal justice with the history of on-site drug testing practice were evaluated. These were all laboratory based on-site drug testing facilities, all focusing on urine as the specimen. There were no non-instrumented drug testing devices being used in an on-site capacity at that time for workplace application. Remember, we always reflect back to workplace. The Executive Summary of that report is summarized here. In general, all 11 facilities attempted to perform on-site drug testing in a reasonable manner. No facility performed drug testing in an overtly inaccurate or false manner. However, without proper quality control and quality assurance programs in place, it is impossible to know whether a facility is producing results that are either false positive or false negative. Several sites had serious flaws in procedures that could have impact on testing results. They are itemized.

Personnel. Several sites had personnel performing drug testing who lacked education in a laboratory science. Most sites did not have documentation of personnel training in specimen handling or performing drug tests.

Without an adequate QC program, which is itemized below, a laboratory with poorly trained personnel is particularly vulnerable to producing errors.

Specimen handling. Many sites were using chain of custody forms that do not fully establish an audit trail to reconstruct the handling of a specimen from collection through analysis, storage and final disposition. Other good forensic procedures, such as sealing of urine containers for transport, were not always being followed.

Security. Security of specimens at several sites was very weak, and it would be difficult for these sites to establish and defend the integrity of their specimens. Many sites were introducing foreign objects into the original specimen container before removing an aliquot for testing. At some sites, employees were actually in the testing area at times when specimens were on a lab bench, in preparation for the testing. These procedures raised serious questions with regard to the confidentiality of drug test results performed at that site.

Standard operating procedure manual. Only a few sites had an SOP that covered all aspects of laboratory operation. Many sites used only the manufacturer's operating manual as an SOP, without specific details for specimen handling. A description of a QC program was not found in most of the SOP manuals at those sites.

On-site testing methods. Each facility was using a commercially available immunoassay method for urine drug testing in a laboratory setting and it appeared that tests were being performed according to the directions provided. Most facilities had validated their methods only by the manufacturer's set up procedure. Most were not sending a random sample of urine, negative urine specimens, to a reference lab for re-test to determine their rate of false negative specimens. The drugs tested varied among facilities and according to the company policy.

Quality control and quality assurance programs. Only three of the 11 sites visited at that time had complete QC programs that included open and blind controls. Most sites were using manufacturer supplied calibrators and blanks as their only QC samples. In other words, what was part of the assay requirements by the manufacturer was the sum total of the QC in that batch.

Policies. All sites were taking some personnel action on presumptive positive drug test results. Whether this policy violated employees or potential employees' rights was beyond the scope of this report. It must be recognized that this action does not conform to sound forensic toxicology practices and may lead to false positive results and false accusations about drug use.

In addition to the Executive Summary findings, there were working group recommendations. They include guidelines should be developed that establish operational consistency and analytical accuracy among on-site drug testing facilities. Such guidelines should include, but not be limited to, criteria for the education, training, demonstration and documentation of skills of on-site laboratory personnel performing the appropriate test procedures. Criteria for the standardization of chain of custody and control procedures for specimens and aliquots of specimens. Criteria for the assurance of the security and integrity of specimens and the security of the data. Criteria for the development of a complete and updated standard operating procedures manual. Criteria for the standardization of the drug test, including cutoff concentrations, blanks and standards and controls to be tested at or near the cutoff concentrations. Testing methods to be used and procedures for the validation of those methods. Criteria for the establishment of quality control and quality assurance programs to cover every step in the drug testing process. Consequences of drug testing in the criminal justice system are no less serious than in the workplace, and an inaccurate or unsubstantiated result should not be permitted. A set of criteria similar to the above guidelines must also be developed for drug testing within the criminal justice system. Ten years later, we see that they have taken that quite seriously and have several model programs, which is why we invited them here to share their experiences with us. A system should be developed for monitoring on-site drug testing facilities to assure compliance with the above guidelines, and that may include inspections.

That's the summary of the report. This report was never released in a formal way to the public, but rather formed operating guidance and recommendations to the Drug Testing Advisory Board at the time. We will continue to request information. It's time to clearly re-evaluate where we are with technologies and systems and processes at this time. That is why we are having this meeting. I'd like to ask you to reflect back on some of the issues that were found to be true at the time in a consensus format and let's try that again, based on where we are today and with today's technology.

DR. KADEHJIAN: I'd like to know why that report was never released.

DR. BUSH: It was a report, as I said early in the beginning, it was written and referred to the Drug Testing Advisory Board, and at that time, there was just no need felt essential to release it to the public. We took our guidance and established then our thoughts and our method of continuing to evaluate technologies in the future. Just as the technologies were different ten years ago, so was our approach to developing information. Today, we have much more opportunity to have public meetings. It wasn't a conscious thing. It's not like we were trying to hide anything. The facts are the facts. They were in the report. It's just that we didn't choose to put out a Federal Register Notice or something like that.

Dr. Willette, the moderator this morning for us.

DR. WILLETTE: Yesterday we heard overviews from a variety of programs that do on-site testing, focusing mainly on the criminal justice arena. This morning's presentations are going to focus on workplace programs, some of which have a longer history than others and some that are evolving. To start off, we have asked Dr. Leo, as he's known in the field, to talk about some of the experiences he has had in doing inspections on workplace on-site programs.

ON-SITE DRUG TESTING LABORATORY EXPERIENCE

DR. KADEHJIAN: As Bob indicated, we did speak yesterday about our experience with the criminal justice system. You heard Jim Baer from the Central District of California and Jay Carver from Washington, D.C. speak about the utility and effectiveness of their on-site testing programs, but in reality, we are supposed to address workplace testing here and the due process requirements are certainly different. The statutory guidance we have, regulatory guidance we have for workplace testing is clearly different. I have been asked to speak today about the experience I have had with oversight of some workplace drug testing programs, on-site programs.

There are really three separate programs through Duo Research we have had experience with. We had actually managed an on-site testing program that Amtrack had put together. I hate to mention the affiliation with Amtrack simply because of all the train accidents that have occurred, many times not their fault and certainly has nothing to do with the quality of our oversight of their drug testing program. They had actually had instruments at several of their stations, in Philadelphia, Washington, and L.A. They actually had their nurses running on-site drug testing. We had experience with that program for many years.

After the Exxon Valdez disaster, Exxon had wanted to establish drug testing internationally, and even though there was not a DOT mandate that they do that, they felt they should have some exposure. We actually developed criteria for Exxon's own medical staff to identify laboratories around the world, to go in and do inspections of those laboratories, basically training them as inspectors. As we think about distributed testing in an on-site testing sense, moving outside of the laboratory, the training issues clearly become important. Again, we have had experience doing that.

Finally, today I'd like to talk about our recent experience with oversight of one nuclear power plant. Under the Nuclear Regulatory Commission, power plants are licensed by the NRC, so they are called NRC licensees. Many people are unaware that there is a federally recognized utility of on-site testing in the NRC program. Many people think the federally regulated workplace does not allow on-site testing and they are familiar with the HHS/DOT type testing programs. The NRC as of 1989 published in the Federal Register their Final Rule did allow NRC licensees to have their own on-site testing program. I'm going to read you some excerpts from some of those Federal Register citations, simply because I think they are instructive for our purpose here today. I think it also is significant that the final regs came out in 1989. This was in June of 1989, so it was after the two Supreme Court decisions on Von Robb and Skinner, so in a sense, these regulations did have the benefit of the Supreme Court's rulings on two workplace testing issues, one for Customs agents in Von Robb and the other, post-accident testing in the railroad. Note that the HHS guidelines produced in 1988 did not have the benefit of those Supreme Court decisions, which laid out due process issues. One of the things that was addressed in these final regs was on-site testing. The Nuclear Regulatory Commission -- these plants had tremendous safety issues to deal with. They recognized the benefit of having very rapid and immediate results.

A key point distinguishing the HHS guidelines, Federally regulated workplace under DOT/HHS type testing, is that those testing programs are not detection programs. Those testing programs are deterrent programs. We are constantly reminded about that. The Nuclear Regulatory Commission has said, no, we are a detection program, so the program is in fact called Fitness for Duty. We are not only trying to deter people from drug use. We want to identify those drug users. One of the things that they talked about was the accuracy of screening tests. Many people think screening tests are here and GC/MS is here, but in reality, the courts and certainly scientists have recognized that screening tests can be here and GC/MS is here, and that is one of my sort of pet arguments.

In the Federal Register in 1989, the NRC was acknowledging some concerns about initial screening tests and immunoassays. I just want to cite from this Federal Register, and they relied on the Federal Aviation Administration's comments, and they said the FAA in their response to comments in November of 1989 provided a clear response regarding initial screening tests, that we find no reason to improve, and in that they said, contemporary screening tests such as immunoassays, have become extremely accurate and approach 99% accuracy levels. This is now the NRC's review of where initial immunoassays stand, and they are saying the FAA says 99% and we see no reason to improve on that.

When they now start drafting regulations to deal with on-site testing, they are starting from this standpoint of having very high regard for these initial screening tests in support of that program.

SPEAKER: You have to realize who was writing those comments, Leo.

DR. KADEHJIAN: If you haven't already gathered, I'm a major supporter of the use of on-site drug testing, whether it be instrumented or non-instrumented systems, with all the appropriate caveats, of course. In the nuclear regulatory industry, many times they have power outages, and as a result, they suddenly need to have many contractors show up, have immediate access to the facility, so they have to screen many people quickly, so they are looking for tests that allow them to do that, and on-site testing is very effective. One NRC licensee said that delays in granting access caused by the loss of pre-access on-site testing could cost us $15 million annually. Again, a recognition of the economic benefit of having rapid results. That was in the Federal Register in 1991, after the 1989 program was put in place.

Finally, in 1996, there were some proposed rules, really quite a large set of proposed rules, but even in that Federal Register, they did address the use of non- instrumented tests within the NRC. That is everything up to that point, although it didn't preclude non-instrumented testing, it certainly was related to instrumented testing. In the 1996 proposed rules, just out of fairness, they said they had received requests from licensees and vendors to permit the on-site use of non-instrumented qualitative immunoassay methods that involve the use of inexpensive devices. In 1996, what they had to say about it is while Part 26, 10 CFR 26, which is the NRC's Fitness for Duty program, it doesn't preclude the use of such devices for screening tests. The NRC is aware that there are several technical variables involved in the use of these devices that may prevent them from achieving the high levels of specificity, accuracy and repeatability demanded in licensees' drug testing programs. They were taking a dim view of the use of non- instrumented tests. They indicated, for example, there are no quality control procedures known to the Commission, Nuclear Regulatory Commission, that could be used to validate the results produced by these devices, given the uncertainties surrounding the potential use of non- instrumented testing devices, the NRC would prefer that these devices not be used for screening tests and licensees' Fitness for Duty programs at this time.

They in 1996 took a dim view of the use of non- instrumented devices in their programs. They did indicate that the NRC is aware that HHS has been mandated to investigate the accuracy and reliability of these devices, and that's effectively what's happening here. With that framework in mind, again, those are proposed rules in 1996 that have not been acted on yet, I'd like to talk briefly about a program that Bob Willette and I through Duo Research operate in overseeing one of the licensee's programs. I won't mention the name of the licensee.

There are 90 some odd nuclear power plants in the United States. I guess about a third of them now do some form of on-site testing. They are allowed under the NRC to use those on-site test results for employee action. Let's be careful about what that means.

If they can demonstrate that they have 85% or higher confirmability of their THC and cocaine results, then they can use those results to remove an employee from a safety sensitive position pending confirmation without any other evidence, only the drug test evidence is sufficient to remove an employee from the safety sensitive position.

An individual tested positive for cocaine, we don't want that person pushing the buttons on the nuclear reactor. Please go the library, write your monthly report, go home, sit by the pool, whatever, we don't want you in that safety sensitive position pending confirmation. No evidence of impairment is required for THC and cocaine.

MR. McCAN (ONDCP): Would it be fair to assume that they also used that to exclude applicants for employment to safety sensitive positions? It's probably a larger group of people than the actual employees which are taken off that status.

DR. KADEHJIAN: I don't intend to be a spokesperson for the Nuclear Regulatory Commission or its regulations. I believe that 10 CFR 26 does not apply at all to pre-employment testing, but it does apply to pre-access, if somebody is going to have unescorted access to a secure area. These regulations don't apply to pre-employment.

A key issue is it is saying you have to demonstrate quality in your program, and the fact the NRC did review what data was known about their on-site testing programs, what was the experience of licensees with on-site testing, what level of confirmability did they observe. They reviewed many nuclear power plants and they found that they had quite high levels of confirmability. For cocaine, it was 90 percent. For marijuana, it was 88.5 percent. They said if you can demonstrate 85 percent confirmability rates for THC and cocaine, you can remove them from the safety sensitive position pending confirmation.

This particular facility that we oversee does not use on-site positives in any way. They are simply sent on for confirmation. They are not used for employee sanctions. Again, their goal is really to identify negatives.

I'm just going to summarize a few aspects of our inspection. We have developed an inspection checklist that's modeled after the national lab certification program, as well as the specific regulatory criteria mentioned in the Federal Register.

We have been involved in this program from 1990. We have done inspections just about every two years from 1990, and those inspections involve not only review of the laboratory and collection testing, on-site testing, confirmation, all of that, but also review of the medical review officer function. We will audit all of the records that the MRO has when they review their programs. We have also been involved in an audit of their confirmation laboratory. That is they are required under their collective bargaining agreement to go into their confirmation lab. I was involved in being one of the members on that audit team of reviewing how the confirmation lab works. They submit specimens to the confirmation lab and I've just outlined the specimens they submit. Any client that tests positive, any employee positive, if you will, get submitted. They also are required to participate in a blind QC program, where they will receive specimens from Duo Research. They are packaged up to look like employee specimens and submitted as blinds to their confirmation lab. They also submit negative QC's. That is the Nuclear Regulatory Commission was concerned about on-site testing missing positive specimens and wanted to assure against that, so you must submit a proportion of your negatives from each day's testing onto the confirmation laboratory.

Also they have decided that the Fitness for Duty staff, that is those who are involved in the Fitness for Duty program, should not be tested by themselves, so all of their specimens, if they are pulled up on random testing, periodic, whatever, they will all be sent to the confirmation lab. They also have some employees who work off site from their regular facility, and those specimens are all sent. These are all the specimens that are sent to the confirmation lab. They actually have two laboratories they are using for confirmation, one local, one further away. They submitted 120 and 175 specimens. There were 24 of those 120 that were employee positives. The rest represented these other specimens. Fourteen of those 24 employees did not confirm at the confirmation lab. For the 23 employee positives sent to the other lab, 8 did not confirm.

I indicated that they needed 85% accuracy on their THC and cocaine confirmations in order to take action, but because they are not taking any action on employee positives, this was not an issue for them.

When I go in to do the inspection, I review every single one of these results. I pull out all the on-site records. I pull out all the confirmation lab records and I pull out the medical review officer's records in reviewing those results. In none of these cases were any deficiencies found. That is there were valid medical reasons for why these specimens did not confirm. Many of the specimens in fact did confirm on repeat immunoassay screening at the confirmation laboratory, in effect, confirming the on-site result, but did not confirm under GC/MS criteria. There were obviously a myriad of reasons for this. Dextromethorphan use causing PCP positives. Border line results for THC that were just below the cutoff on confirmation. Again, valid medical reasons. No deficiencies were found for any of these.

The QC program. They are required to submit a sampling of negatives. They have chosen to submit one out of every 50 specimens, so lab one got 55; lab two got 55; every single one of the negatives was in fact correct.

The blind QC program, NRC regs require at least 10% of all of the specimens submitted to confirmation labs, at least 10% must be the blind QC specimens. Of those specimens, there were 33 out of 120 that were blind QC's and 37 out of 175 were blind QC's, to the two separate laboratories. They were clearly far exceeding the NRC requirement for sending in blind QC's. Again, every blind QC was correct, on-site, by their automated immunoassay and they are using an Abbott Accent Analyzer. They had two ADX's before that.

All of those QC's were correct on-site and they were also all correct at the confirmation lab.

DR. SMITH: I would like to ask a couple of questions because I am somewhat confused. Number one, if you could go back to looking at the fails to confirm, from the 24 and the 14. You made a comment that this could be due to valid medical reasons, which would seem to me that you are talking about the MRO verification to negative, which has nothing to do, I don't think, with this. Is that correct or incorrect?

DR. KADEHJIAN: For example, there may be hydrocodone in the specimen and maybe the confirmation lab isn't testing for hydrocodone, the person has a valid script. The initial test is correct. It identified the hydrocodone. The confirmation lab report being negative is correct, and the MRO's review of that is correct. There is no deficiency even though it did not confirm at the confirmation lab. That's sort of analogous.

DR. SMITH: Again, do you have any data that shows of those 24, when 14 did not confirm, what percentage of those were, for example, for the opiate class of drugs or the amphetamine class of drugs, where we typically have seen much more cross reactivity with substances that ultimately are not the analyte for GC/MS confirmation or were these in fact also cocaine or marijuana THC failure to confirm?

DR. KADEHJIAN: When I do my MRO function, I review every single one of the results that they look at. Every specimen that was sent to the confirmation labs was examined and their results, and all MRO results were examined. I know what every single one of these 24 specimens was, what it tested positive for, what the MRO review was.

DR. SMITH: Why would a screen positive specimen but a confirm negative specimen go for MRO review? I'm confused.

DR. KADEHJIAN: Because the MRO has oversight of the Fitness for Duty program and on-site positives are reviewed by the MRO. Think about a SAMHSA lab that gets a positive that doesn't confirm, good QC of your laboratory procedure might say why did that happen, are we making mistakes. You need to take a look at why that happens as part of good QC. They do that. The MRO reviews those on-site positives. They are allowed under the NRC regulations, by the way, to have access to all of that information, but it's limited to the Fitness for Duty staff, the people that do the testing, the MRO and the EAP as necessary, but not management. Management does not have access to these on-site positive results.

DR. SMITH: They do not have access?

DR. KADEHJIAN: Right.

DR. SMITH: That was my question also. You had stated earlier that in this case, let's say somebody has a positive for cocaine. Fitness for Duty pre-access test. That positive result, that person is going to have access that day anyway; is that correct?

DR. KADEHJIAN: That's correct, because they have chosen, this licensee has chosen not to act on on-site positives but the MRO has access to the information.

DR. SMITH: And contacts the employee?

DR. KADEHJIAN: Yes. Management actions are not taken against that employee. They are not removed from the safety sensitive function, even though under NRC regs, they are allowed to do that.

DR. WALSH: You have indicated 24 employee positives, are any of those blank QC's?

DR. KADEHJIAN: No.

DR. WALSH: The reality is there are really only 87 if you subtract the 33 blind QC's, there are only 87 employee specimens and 24 of them are positive? It's not a very good screening system, it seems to me, if a third of them are going to be positive and then 50% of the positives are not going to confirm.

DR. KADEHJIAN: Realize that this site tests several thousand specimens a year. Their rate of positives is like this, very few employees test positive. One of the summary statements I want to make is that I again have been very impressed by the way this facility runs its program and that everybody involved who I've had the chance to meet and talk to in the nuclear power industry is exquisitely concerned about safety. The rate of drug use is really very, very low. Out of all the thousands of specimens they test, they had only 47 employees who ever tested positive.

I'm just showing you the numbers. There were 295 specimens that ended up going to the lab for repeat testing, confirmation testing, or the negative QC's, the blind QC program. Out of all those specimens, only 47 were employee positives. I'm just saying here are the results that they had. The key point here -- this seems like a big number, that 14 and 8 did not confirm, what's going on, are they making mistakes here. It's just the realities of testing. No deficiencies.

DR. WALSH: What I'm suggesting is that the technology they are using doesn't seem to be very good, in terms of screening, effectively screening out negatives, true negatives from presumptive positives.

DR. KADEHJIAN: I'm not sure how you make that statement from this.

MR. SMITH: These that screened positive at the site and then screened positive by the lab and then were negative by GC/MS, were they checked for nitrite?

DR. KADEHJIAN: No, I do not believe there was nitrite analysis done at this time.

MR. SMITH: The nitrite, of course, as you know, would cause a GC/MS screen to be negative with the other two being positive.

DR. KADEHJIAN: I do have creatinine, specific gravity, and pH data, but I don't believe any of these were looked at for nitrite. I don't want to belabor the point here. I'm just trying to give you a picture of the program. I think more importantly is the diligence with which they carry out their on-site program. Again, no deficiencies were found in any of the testing that was done at this program.

DR. SALAMONE: The cutoffs are probably lower than what's normally used, right? These are SAMHSA cutoffs?

DR. KADEHJIAN: They are SAMHSA cutoffs. Also note that the NRC does allow their licensees to test for other drugs, so in fact they were testing for barbs and benzo's. They are discontinuing that, but they are allowed to do that under NRC criteria.

Finally, I'd like to just indicate that as part of my MRO review, I ensure that the MRO has the appropriate education, training, experience. That is, they are knowledgeable of substance abuse issues. I will query them about all of the interpretative issues that we are now aware of. I review all of the procedures that the MRO goes through, pull all the records out, make sure they have appropriate chain of custody and review, as I said, all of their results.

It's generally a two day inspection to do both the laboratory and the MRO function. Just to remind you of the inspection checklist that we use when we do this, again we have basic information about the collection site. We go through all of the collection related issues. The Fitness for Duty staff are registered nurses. They have been through several days of training on numerous occasions on the instrument, go through all of their ability to interpret everything that the instrument prints out. They have tremendous security, chain of custody. They take three specimens. They take an aliquot and two reserves. The aliquot is for on-site testing. One split is a reserve. The other split goes to the confirmation laboratory.

The diligence with which they approach all of this is really dramatic in my view. They bend over backwards to do everything right. I'm very supportive of the role of this program and its success. I think they do a great job.

DR. WILLETTE: We will have time at the end of the morning session for questions and answers, and maybe to make sure we stay on schedule and that we get all of our industry speakers in this afternoon, we will get out of here at a reasonable time with flights, maybe we can hold off until the end of each presentation, if there is a little time available for quick questions, we can do that.

Our next speaker has been very generous in providing the Board with updates on a large scale on-site program that's going on with the U.S. Postal Service. Julie Murdoch with Bensinger DuPont & Associates is here today to give us a further update and maybe a little more insight into how that program is going.

U.S. POSTAL SERVICE PILOT PROJECT UPDATE

MS. MURDOCH: Bensinger DuPont did a study. There were only three or four products that were available that were FDA cleared, that was an important point that the Postal Service wanted, and that were available for use, so they did a study -- we did a study, and then I came on board.

The decision was made for a variety of reasons to use the TesTcup for a pilot project manufactured by Roche. Probably the biggest reason why is they did very well on the reliability study, but also there had been concerns raised about collectors having to manipulate urine, having to drip or to dip or somehow handle the urine, so the cup device was selected.

During 1997, we developed the protocol for conducting the on-site testing program. Bensinger DuPont took the lead on that, but we worked very closely with the Postal Service in assessing what their needs were, how they thought things would operate in their contract collection sites as well as their medical units. We worked with the Postal Service's Legal Department, so every step of the way, people were blessing the product that we were coming up with. I'll talk about the protocol in a minute.

We also at the same time continued to assess the TesTcup in particular. It was subjected to a variety of different studies, which the people who are here from Roche may remember, grimly, including spiked specimens, clinical specimens, and one of the things we did was we took our draft protocol and the TesTcup and we put it out in four collection sites, and we ran a side by side study where we collected specimens from individuals and the donors did know that their specimen was going to be used for this study, the TesTcup analysis was not performed in front of the individual, but they were informed that a portion of their urine would be used for that purpose, and they did have the right -- donors did have to sign a consent, an informed consent, obviously, and they did have the right to opt out of the study if they didn't want that to happen. We did the analysis over a couple of months. I guess we eventually ended up with about 900 side by side results, which were very satisfactory.

We finalized the protocol and in August of 1997, we conducted training via the Postal Service television network, which was very cool. They have their own television station. I don't know if you all know this, out in Potomac. They broadcast by satellite link to 90 sites. There are about 5,000 people who are on the other end, which I fortunately didn't know at the time I was doing the training, and we had pre-positioned for all of the actual people who were being trained to be collectors training materials. They had the outline that we were using. They had the written protocols. They had the information for donors which we had drafted. They had specimens which they were actually going to be using for practice and TesTcups obviously, and then they had an examination, unknown specimens for their examination.

We did the training and the examinations came in by the hundreds, and I got to grade them all. Those who did not pass the initial examination were subjected to retraining and they had to be re-examined. There were a fair number of people who didn't pass the first time. Until somebody had passed the Postal Service examination, both the practical aspect and the written training examination, they were not allowed to conduct Postal Service on-site tests, and that remains the case today, as a matter of fact.

Over the next couple of months, up until about January of 1998, we did two things. One, we rolled out the pilot contract with Roche and started pre-positioning supplies, which when you are talking the volume that we were talking, it was fairly substantial.

The program was actually rolled out district by district. There are about 158 districts in the United States, so we had a little bit of lag time between when different districts rolled out. We also had to re-negotiate with the Postal Service Laboratories, who were not happy, as you might imagine, because their volume was going to decrease substantially with their number of positives, their rate of positives was going to increase. The Postal Service contracts had been let on a per test flat fee basis, and obviously, since that had been negotiated based on a specified rate of positives versus negatives, that just wouldn't be fair to require them to stick with that cost. That was re-negotiated and the Postal Service went to unbundled pricing with the laboratories.

All of this was sort of going on as a parallel track. In addition to all that, the first tier of trainers, of people who had been trained, which were the occupational health nurse administrators in each district and their assistants, were going out and training collectors at different collection sites, and I was getting all those exams in. Fortunately, I don't have to do that any more. The district nurses do all the grading.

We had second tier training going on, and we had developed a train the trainer program for the occupational health nurse administrators, so they have overhead's which they have to use and a script which they have to use to do the training, so everything is consistent at every site where training occurs. They also have to use the same specimen sets for the examination. We were really pushing for consistency. Throughout 1998, the program rolled out and it was a whole lot more successful than I think any of us actually expected it was going to be and a great deal less painful than I think any of us expected. I'll talk about some of the lessons learned.

In January of 1999, the Postal Service decided to make what had been a pilot project a full time program of the U.S. Postal Service and they bid out the program. The Postal Service received 11 offerors submitting 13 products. The bids went through full evaluation by both a laboratory evaluation of the devices. Each offeror had to submit 50 of the devices. Laboratory evaluation was done and also a technical evaluation was done by a team of Postal Service individuals. It was objective. It wasn't based on past practice, but Roche won. They were awarded the contract in March of 1999 and they are the current provider. The contract includes, by the way, not just the testing device, but also collection cups, which the Postal Service has specifications for, and Adulta Check strip. That is a bundled service that Roche provides.

To date, Jim Baer mentioned this yesterday, one of the problems with going to on-site testing is the fact that it's difficult to track data. If it's laboratory testing, you get the data from the laboratories. You know how many specimens went in, how many were rejected, how many tested, et cetera. With on-site testing currently, the Postal Service does not have an automated database. They can only guess how many on-site tests have actually been conducted to date.

About half a million TesTcups and ancillary products have been sold and our best estimate is between 300,000 and 350,000 tests have been conducted using the on-site device, which is, I think, a reasonable number to assess what lessons have been learned.

Probably the first and biggest lesson that we learned is that our protocol did work. One of the big concerns that you hear people mention, you are going to have confrontation, you are going to have donors who are going to get upset because of their drug test results, et cetera. I am the point of contact in every training that goes on and with every piece of written literature that goes out. I'm the point of contact for problems having to do with this program. In two years, I have never received a complaint about an upset donor. Never. I've received lots of other complaints, but I have never received that particular complaint. Believe me, had it happened, I would know. I would probably still be paying for it. No. It hasn't happened. That was our guess initially. That's why we set up the protocol the way that we did, and to go through the protocol very simply, the way it works is the donor walks in, shows identification, just like in the current collection process. The specimen is collected using a separate specimen container that has a snap tight lid attached to it. The specimen is checked like usual for temperature and quantity and that sort of thing. If it's out of temperature range or it looks peculiar, the individual can give another specimen. At that point, 30 milliliters is poured off into a TesTcup that has just been opened, taken out of the pouch. The Adulta Check is run at that point in the TesTcup. If the Adulta Check shows an abnormality, we never say positive. If it shows an abnormality, that's the end of the testing on that specimen.

The reserved portion, which has been in this nice leak resistant collection cup sitting right in front of the individual, is then poured into a specimen bottle, sealed, put under chain of custody and sent to the laboratory. The chain of custody form is annotated with whatever the abnormality was and we have protocols for the laboratories to assess that. If the Adulta Check is okay, then the TesTcup is run. If any of the TesTcup results indicate the possible presence of drugs, we call that "further testing needed." Never positive, because it's not. I don't care what you say, manufacturers, it's not positive. We call it "further testing needed" because that's what it is. Again, that's the end of that specimen. The reserve specimen that has been sitting there in front of the donor and the collector the entire time is then taken from that collection cup, poured into a separate specimen bottle, sealed, put under chain of custody and sent to the laboratory for screening and confirmation, and the laboratory is not told what drug was identified in the TesTcup. They just go through basic screening and confirmation.

We maintain a log book. Specimens are not put under chain of custody until custody changes. Custody doesn't change if the donor is sitting there and the collector is sitting there. The tests are recorded in a log book. Only if the specimen goes to the laboratory is a chain of custody form begun, and that chain of custody number goes in the log book as well. The results are all annotated in the log book and there are initials for the donor and initials for the collector. It's fully forensically satisfactory, but it's not a whole lot of pieces of paper, which was one of our big goals with the Postal Service, to reduce the volume of paper that the medical units were having to handle, because you can imagine with 220,000 to 250,000 pre-employment tests a year. That doesn't include all the other stuff that they do.

That's the way the protocol has worked. It has worked remarkably well. As I said, we have not gotten any complaints about donors confronting individuals or collectors. I received calls from collectors, from occupational health nurses, who have had instances where people walk in, find out they are going to be subjected to an on-site drug test and say how maybe they don't really want to work for the Postal Service and they turn around and leave, which is not a bad outcome, all things considered. There is something about that immediacy of the result that seems to be producing more of that than the regular drug testing process.

The Postal Service has done pre-employment drug testing of its applicants since 1986, something like that, so a long time. This is a relatively new thing. That is not to say that the implementation has been without its difficulties.

One of the biggest issues that we addressed was the fact that on-site drug testing is anathema in some states or some state laws say it comes under their state clear requirements or that the collection site has to have certain certifications and qualifications, none of which fortunately apply to the Postal Service because it's a Federal agency, and they get preemption, which is one of the nice things about having the Postal Service as a client.

We were able to implement a program nationwide and see how it works, but dealing with the state health administrations and explaining to them that they couldn't get involved in Postal Service business was difficult, and a lot of the contract collection sites wanted an affirmative writing from the Postal Service to that effect, so the Postal Service Legal Office did a legal analysis of preemption issues and basically told the states to get out of our face, you know, with lots of citations and well put. That's essentially what the memoranda came down to, that the Postal Service had Federal grant of authority and didn't have to comply with these particular regulations. That memorandum is made available to state health authorities if they raise a concern about the Postal Service program. We had a number of those early on in the program. I haven't gotten any requests for it in about six months. I'm hoping that issue has largely gone away.

For those of you who have a non-regulated private client, obviously you are going to have to be aware of those requirements and deal with them. That can make it difficult for multi-state employers. Just because you are doing the same thing in each state and you think that's equitable, it doesn't mean that it's legal. They are two different things.

We had a lot of concern from contract collection sites about liability and that was mentioned yesterday. I think Jim mentioned that. They were concerned first about the confrontation issue, would the Postal Service indemnify them in the case that somebody -- I'm not going to use the term "went Postal," somebody got upset during the collection process and harmed one of their collectors. Of course, those are all standard provisions in Postal Service contracts. It's a Federal contract. That was a concern. The possibility of mis-reading results and harm occurring, false negatives, false positives, that sort of thing.

Those were all issues that had to be addressed, and that gets back to why we spent so much time validating the device up front. The Postal Service did not want to institute a program where people were going to be put to work in Postal Service facilities based on an on-site negative who would have been precluded from going to work had the laboratory program continued. What if they caused harm, and the question was asked, did the Postal Service maintain a particular standard of care. Was the Postal Service acting in a reasonable manner when it used on-site testing as opposed to laboratory based testing. We spent a lot of time up front making sure that we could say yes, this was reasonable.

Some of the performance issues, one of the things that we have done since the program went into place is track any unconfirmed specimens that went to the laboratory listed as "further testing needed." We got a whole bunch of them right at the beginning of the roll out, primarily from two or three sites. What was identified, we actually went back and looked at many of the data pertaining to the specimens. Some of them were re-run on the TesTcup to see what happened. A large number of them were just negative, especially from one particular site that I think was responsible for most of these unconfirmed FTN's.

What we eventually found out was happening was the collectors were taking a very conservative approach. If they couldn't absolutely call it negative, they sent it to the laboratory, which is not the way the program is supposed to work. It is supposed to work if it looks like it may be negative, it's a negative. Any color in the field is a negative. That's the way the device is set up. If it wasn't a crystal clear line, they were sending it to the laboratories, so we had a lot that were coming back. There just wasn't any drug there. Some were cross reacting and some were below threshold on the confirmation. We had a concern raised especially from one site about whether specimens, because they were going for both screening and confirmation, whether we were losing analyte and we were having specimens come in below cutoff that would have been confirmed had they just gone for GC/MS. That turned out to be not true. I think in all the ones that we looked at, there were maybe two or three that were right at the cutoff and who knows whether you can contribute that particular level to a decline over time or the TesTcup just happened to be more sensitive, who knows, but it did not seem to be a problem.

I also get the leaking specimen cups, the cups that have -- we had one batch where the lines only came halfway through the window, like in the test valid. The lines only came halfway through each window. It turns out in the manufacturing process, the reagent strips were just a little bit off center, so the reaction line just went halfway through each window. They were perfectly appropriate testing strips. They were working fine, but they looked weird, so we dealt with the weird ones. That was one small batch and that has not been a problem since then. We also had a problem with creatinine on the adulteration strip with one batch that seemed to resolve itself relatively quickly.

What are some of the elements that I think are critical on establishing an on-site testing program based on what the Postal Service has gone through over the last three years. Advanced planning is a big one. If you are going to roll out a program of this size, you can't do it overnight, by any means. You have to start with validating the product that you are selecting, for one thing.

Manufacturers, you can close your ears for a minute. I know they say their products are FDA approved. I know that. I've seen all the data. FDA approval means these particular studies that they ran, God knows how many other ones they ran, these particular results were sent to the FDA for review. It does not mean that an actual practice in new batches and in new lots with perhaps different manufacturing processes, you know, slightly refined as chemists want to do, that the same results would be obtained. Production runs are different from FDA pre-market approval or 510(k) clearance runs. That's been my view. The truth is what the manufacturers put in their sales literature is sales literature, and if you rely on that, you are just like somebody who walks into a car dealer and relies on the car dealer's representations about this used car, this car, without taking it out and driving around. It's just not sensible, I don't think. You need to validate the product using whatever your own methodology is, whether you do it by doing a review of literature applicable to that particular product, by doing your own validation studies, however you do that, make sure that you are not just looking at one manufacturer's representations of one or two limited studies of the best possible information, non-production run devices sent to the FDA.

Have a written protocol. This cannot be done off the cuff. Not only are you going to have problems with training collectors because you have to train people to something. If you have a written protocol that covers all of the elements that might come up during a collection and a step by step procedure for people to follow no matter where they are and whom they are testing, you reduce your possibility of error. You reduce your possibility of legal exposure and there are not due process issues. You reduce your possibility that one collector is going to do testing in one completely different way than a collector who is in another location. It gives you a more defensible posture if you can say this test was conducted using these procedures following these checklists, and that's also what you train your collectors to do.

I know these devices are relatively simple when you look at them, but they are actually not that easy to use once you put them in the context of an entire testing protocol, once you are not just looking at it sitting here playing with it in a conference, but in a workplace setting, in a collection site, where you have to consider not just getting this result on this person, but making sure the paperwork is done, you have the right person, you have the right company, you have the right pieces of paper, whatever.

You have to have training. Actually, some of these devices, the reading of them is not particularly intuitive. I love the pictures in the manufacturers' brochures, this perfect clear pink or blue lines, or absolutely white windows for a presumptive positive. That's not the way it works in real life. Somebody talked about lighting. One of the informal pieces of guidance that we have given our collectors is if you can't read it, take it over to a window because the blue lines seem to be a whole lot more discernible in natural light than they are in fluorescent light.

People need to be trained. You get iffy results and people need to know, your collectors need to know how those should be read.

Monitor implementation. We do a quarterly QC of the TesTcup and the AdultaCheck. We do monitor the unconfirmed FTN's, although probably we are going to terminate that because I'm just not getting very many reports any more. Considering the volume of testing that's going on, it's below what we would have expected for normal variation or cross reacting substances. We are probably going to terminate that. It's been two years.

We do have in the QC program a negative QC protocol for having negative specimens sent to the laboratory. Unfortunately, the Postal Service is in a budget crunch, believe it or not, and they have not funded that yet. At some point, that will be brought into play.

The Postal Service does inspections of its collection sites. Its occupational health nurse administrators go out on-site to the different contract collection sites and the area occupational health nurse administrators go to the district sites and then we do the areas, when it's funded.

What don't I think is necessary based on this experience? This is me speaking personally. We are really not talking Postal Service now. I don't think donor anonymity is required. I don't see any reason for it. It has worked very well to have the collector and analyst be the same individual. We don't have to worry about things like chain of custody. We don't have to worry about things like the temperature of the specimen declining, which affects the sensitivity of many of these devices, so we don't have to worry about false negatives going out because we have cooler specimens. Everything happens in front of the donor. We have donor buy in into the process. That has been important. We have never gotten challenged, never, not in court and not informally. Surprisingly, once the collector is used to it, they kind of like it. It's fast. They get to see something from start to finish. They reduce the amount of paperwork. They don't have these chain of custody forms everywhere and they seem to like it. The assertions that if you don't have anonymity, you don't have a defensible program, I don't think that's true, and that it may somehow introduce some sort of possibility of conflict of interest or bias, that's no different than if you have a collector here and an analyst who happens to see his friend Joe walk in or knows his friend is coming in for a test and can identify that was his specimen and he's going to falsify the results.

Life's a difficult situation now and then. You just have to be willing to balance the costs and the benefits, but costs associated with obtaining donor anonymity to me are substantial for almost no benefit that I have seen.

We heard yesterday a very good assessment of why a second analyst is not necessary. It doesn't appear to add anything to the process. We have done that in our own studies. We have had multiple reads of the same specimen and have had concurrent results across the board, with the exception of one or two. I don't see any reason for MRO review of negative results, negative on-site results. I'm not talking about the ones that aren't confirmed by the laboratory. I do get calls by the MRO's occasionally on those. I don't think there is any necessity for chain of custody on all specimens if the testing occurs in front of the donor, because there is no change in custody. You don't have to have a chain of custody to reflect that. Chain of custody is legal fiction. It's a piece of paper substituting for changes in custody. If there is no change, you don't need a piece of paper. We talked a little bit yesterday about Federal site and collector or analyst certification and inspection. I strongly believe that the Federal Government should establish standards for collectors and for where this type of testing can occur. I'm just not sure it's feasible for the Federal government to actually institute a certification and inspection program.

The Postal Service currently has about 1,000 collection sites doing the testing for them. Eventually, that will get up to about 1,250 when everything is rolled out. There is still one area that is very slow. That is just one company. I don't see how it is possible to look at the 20,000 or 30,000 or 40,000 collection sites extant now and actually do an effective inspection program. It's difficult for the Federal government to keep up with 70 laboratories. I'm not sure how they could do it with on-site collection sites. Fundamentally, I think what we have discovered is that when you are establishing an on-site testing program, you have to do so looking at the reality of what on-site testing is. You can't impose a whole lot of burdens on the program. You can't involve too many people, for example, the second or third review of the result, or you make the program unworkable. You have to get a balance between ensuring you have enough control over the establishment of the program and weighing it down with so many restrictions and burdens that it falls its own way.

DR. ISENSCHMID (DTAB Board Member): You talked about the testing of the drugs, but since you are also doing the adulteration checking, how is that working for the program and what is the concordance with that and what is the Postal Service doing with those issues?

MS. MURDOCH: Surprisingly, there are few adulterated specimens in the on-site program. We had a period of time when the Postal Service was not using the creatinine test on the AdultaCheck, so I don't know during that period what the results might have been for dilutes. We have had, I think, one glutaraldehyde that I've heard reported and a few nitrites, but very few, and the laboratory looks to see whether those are confirmed or not confirmed. By and large, they have not been confirmed, and I have a feeling that maybe because of the fact of what's being picked up by the dip is not necessarily nitrites. I haven't heard of one recently but it may have been pyridine, for example. It will look like a nitrite read on the dip, but if the laboratory wasn't testing for it, it wouldn't come up positive for nitrites. Now that all the good laboratories are going to full panel adulteration testing, we will see more of those, I think.

DR. NIEDBALA (STC): You went through a lot of statistics in terms of the number of trainers and train the trainers and ultimately at the end, you said there is 1,000 sites and that will go up to about 1,250. I can appreciate that after having done the alcohol model. A couple of things. Way back in the beginning you said about 30 percent of the people coming in to be hired went away, didn't get the job or whatever when they had to send out the sample to the lab. What are the stats now? Was that 30 percent totally due to the fact that they had to wait on their drug test?

MS. MURDOCH: No. As I said, there were a number of factors that the Postal Service was looking at. The time required for the medical testing. They do eyes and ears and all kinds of other stuff. They do background testing for people who are going to be actually handling mail. They do a number of things. For all of those, they were trying to reduce the turn around time to do that, and drug testing was one of them. It seemed to be across the board, across all the areas that the last thing that was being waited on over and over again was the drug test result. I don't have any statistics on how much that has been reduced. However, I do know that the human resources people, the transportation management people, the folks who move the mail, who were pushing for this program initially, are thrilled. They have at least anecdotally told Headquarters, which is why the program was approved for full scale implementation, that it has substantially reduced their recruitment costs.

DR. NIEDBALA: Although you didn't mention it today, but I have seen it before, you said that this program actually cost more for them than did the lab based testing.

MS. MURDOCH: Yes, it does.

DR. NIEDBALA: They still feel good about that?

MS. MURDOCH: Yes. The direct costs of the testing between the two laboratories averages about $9 a test and the cost of the on-site program runs about $10.60, $11, something like that. I'm not sure of the exact figure. It's a little bit more. When you spread that over a lot of tests, it's a big number. Also, some of the collection sites asked for a higher per test fee for doing the on-site analysis, and some of them got it and some of them didn't, because those contracts are negotiated locally. Interestingly, some of the districts said you want to do Postal Service testing, you will do Postal Service testing at the current rate. They either accepted it or they bid it out to another collection site that would give them the same rate doing on-site testing. Other districts re-negotiated, but not for a whole lot higher. We had some come in that wanted to double the cost. I would get frantic phone calls from a district occupational health nurse saying they are going to cut off our contracts because we don't have this in the budget. The answer was find another collection site. That was easy, so they did that. Actually, that's an interesting organizational concern because the money for the testing program comes out of the medical budget, but the recruiting costs savings applies to the human resources and personnel and the operational side. Let me tell you, there are some unhappy people over on the medical side until they started getting lots of pats on the back and maybe we can beef up your budget from the operational side, and that's what has happened.

DR. McCOY: I'm interested in your comment about no need for chain of custody, do you have written documentation that the employer or applicant gives permission for the test and do you have documentation that the collector and tester perform the test?

MS. MURDOCH: There are actually three pieces of paper. The first is anybody who applies for a job with the Postal Service has to execute a consent to undergo testing. There is also a background check they have to agree to. There's an affirmative consent to that. They also have the qualification for hiring, personnel notification form. That is the little check block that goes back to the human resources people when they are done with not only their drug test but also their medical assessment, and they are either qualified or not qualified and the basis for the lack of qualification or pending. That goes back to human resources or personnel, and then there is the log book that tracks all of the tests, which are tracked by collection site.

MR. EVANS: You mentioned there had been no legal challenges to your program. This has been our experience also with on-site testing. There has been some challenges in the Federal courts and state courts and criminal justice on-site testing has been upheld in every case. To what do you attribute the lack of legal challenge with your program?

MS. MURDOCH: That's a good question. I suspect, who knows what lawyer is going to find a good job tomorrow, but I suspect there are two reasons. One, the Postal Service has been doing drug testing in the pre-employment context for a decade, more than a decade. It's not an unknown in whatever community that this is a requirement. Second, I think more important is the fact that when individuals come in to be tested, there's a process that's very formalized. They are given a piece of paper that tells them what's expected of them and what's going to happen. There is a five page document that if they want more information about it, each collection site is required to maintain, and they can read every little detail about what is going to happen. There is a checklist for the collectors to use, and I think the fact that the donor gets to see the test result. There are no secrets. It doesn't go back into some secret room and magic occurs and all of a sudden the person is disqualified for hiring. I think that is a big part of it. Also the fact that in every testing situation that is a potential problem, the individual is advised by another piece of paper actually, that the specimen has been identified as needing further testing, but that no action will be taken until a laboratory confirmation is received. Maybe it's just the volume of paper that just shocks people so much that they think it has to be sustainable. I really think it's having all our cards on the table. Everybody knows what is expected, what is going to happen and what the outcome will be at each stage of the process.

DR. McCOY: The other question is yesterday I mentioned the Alaska on-site testing statute where Alaska requires an hour of training of alcohol abuse and an hour of training in drug testing, drug abuse, but then they also require hours of training in just how to administer the on- site test.

How much time did you need to train people to where you felt they were confident? That's one of the issues that has been bantered around. How much time do we have to train people?

MS. MURDOCH: Our program is actually designed, the whole training program is designed to be done in two to three hours, depending on how long people play with the little specimens before they actually turn to the examination. The oral presentation part, the push part, if you will, the lecture, without any questions, runs almost an hour. Then there is time for actually handling the device, playing with it, using the specimens that have been sent and practicing using the device itself, and the dips, watching the instructors. We have an instructor set that has a low pH specimen and a soaked specimen, that sort of thing. That takes a bit of time. The examination runs about 45 minutes to an hour. After one hour, they have to stop taking the exam.

DR. McCOY: Do you have an education program and an 800 number where people can call if they continue to have questions and get technical help?

MS. MURDOCH: Yes. My number is on every piece of training material, unfortunately.

DR. BUSH: Julie, I have a comment that I must make for the record. You mentioned several times that the U.S. Postal Service was a Federal agency. It is not a Federal agency.

MS. MURDOCH: Yes, it is.

DR. BUSH: No, it's not, because it is not covered under Executive Order 12564 nor is it covered under Public Law. They do not have a certified drug free workplace plan in the hands of our office. They are not a Federal agency. They are a quasi-Federal agency and do not fall under any of our regulations.

MS. MURDOCH: They actually are a Federal agency, if you look in Title 26 of the U.S. Code, it defines them. They are a wholly owned corporation of the Federal government, but they are a Federal agency for among other things Title V protections, but they are not included in the HHS guidelines. That is true. They do have Federal preemptive authority. They are like a number of other agencies, but they are a Federal agency. Their people are Federal employees. They are covered by FECA, Federal Employee Compensation Act and Federal Tort Reform Act, all of that stuff covers. They are a Federal agency for many purposes, but you are right, they do not come under HHS guidelines, something for which I am sometimes really happy and sometimes just drives me nuts.

DR. BUSH: I needed to make that very clear for the record because we have not yet embraced any pilot programs for any alternative specimen and matrices testing for drug free workplace programs. That needs to be real clear. However they argue their case concerning collection of specimens and performing the on-site testing in states who require certain things, we have no knowledge of any of that, just so you know.

DR. SALAMONE: I think your points about training are correct and evaluation of a particular manufacturer's product is absolutely correct, especially when you are rolling out a program as large as the program that was rolled out. I will mention something about data contained in the package inserts. The new FDA designed control guidelines that state in a very strong way that data can't be picked and chosen and put into the package insert. Everything, including research books, are now open to the FDA for inspection, and if there is a change in process which causes a change in the clinical effectiveness of the product, the FDA has to be identified with it. That is not to say in marketing literature, yes, things are picked and chosen, but clearly the package insert is a very good guidance document and any changes that occur do have to go through the FDA.

MS. MURDOCH: That's true. One of the things you just mentioned was that the FDA standards for the information are new, so of course, anybody who was FDA cleared before these new standards came into place, I assume they didn't have to go back and re-submit.

DR. SALAMONE: They are liable, any changes that have to be done have to fall under design control now.

MS. MURDOCH: If you don't change it, you don't have to put any new information before the FDA.

DR. SALAMONE: Right, and then contained in the package insert is the same information.

DR. OKORODUDU: I think as Julie mentioned, the product is FDA approved. It does not take the responsibility out of the end user. My second comment has to do with the comments yesterday referring to point of care testing and tests according to CLIA. If a product is approved by FDA for point of care, it doesn't make it cleared for testing under CLIA. From the products you are using right now, do you have any documentation about defect rates? Was your examples because of defects in the cups?

MS. MURDOCH: The number of defects that we had has been very small. I would say overall, I've gotten reports of maybe two dozen. Roche was directly contacted on the misplaced reagent strips initially, but that wasn't a very large number of devices. I don't know how many. Perhaps maybe ten that they have gotten contacted on directly where there was either a no test valid line or the specimen leaked. One thing that we did find with the TesTcup leaking, in almost every case, it was because the collector was not putting the lid on correctly. As soon as we retrained them on how to use the device correctly, and that was just a matter of a phone call and their trainer talking to them or their Roche distributor talking to them, the problem went away. I think I have maybe gotten two or three where they were bona fide leaking, and that was early in the program, which is nice to see. You hope that's what is going to happen, since people are handling urine specimens. I don't have a solid number on the rates. It's so small, I'm not sure the rate would be meaningful.

DR. WALSH: I apologize if I dozed off during this part, but the critical point of where the rubber meets the road is when the operator performs the test in front of the donor. Can you describe that in a little more detail, especially what happens when they read a positive result?

MS. MURDOCH: The way that they are trained to do it and the way that it actually seems to be occurring, which is not necessarily the same thing, of course, is the individual gives the specimen to the collector and the first thing that has to happen is the specimen is assessed for quantity and all the rest of it. They are told that it's an adequate specimen, and then the pour occurs.

Excuse me. One of the reasons why we wanted a specimen container with a lid that snapped on was so that you didn't get specimens leaking or being knocked off the table or accidentally tipped over on somebody's desk or whatever. Many of these nurses have desks in the collection site, which is not my idea of a good time, but that's the way they work.

They open the TesTcup package in front of the individual, after checking that the expiration date is not exceeded, and the Postal Service contract requires that 180 days remain on the usable life of the product when it's shipped. They check the expiration date. They open it. They put the cut out, remove the top, pour off 30 milliliters into the specimen cup, and then they open the AdultaCheck bottle, which is required to be closed between each removal of the stick, dip the stick, do the read on the AdultaCheck. That is annotated in the log book whether it's all normal or abnormal. If that's normal, then the actual specimen -- the lid goes on the specimen and it's turned. It takes about a minute for the read to come up. They are instructed to show the result to the individual before they record it in the log book. They record it in the log book and then the individual is asked to initial the log book. They are told in writing in advance that initialing the log book does not indicate that they agree with the results, but simply that the results were shown to them and that the results written down reflect what was on the TesTcup. If they don't want to initial the result, the testing is canceled and that's the end of the process, and they cannot work for the Postal Service. That's it. If they initial the results and it is an abnormal result, an FTN, then in front of them, the TesTcup is set aside and the reserve specimen is taken, now uncapped, poured into the specimen bottle, which has been sealed and unsealed, and then they put the label over, just like a regular chain of custody. One of the things that is absolutely required is that they have to show the result to the donor and have the donor initial in the log book.

SPEAKER: They interpret the results for them?

MS. MURDOCH: Yes. All they tell them is it's FTN in this window for this drug, that's all, or they tell them it's all negative. They are allowed to make a conclusion statement that it's negative because that's the way the Postal Service protocol is established, but they cannot say that it's positive.

MR. LINEWEBER: Julie, is this pre-employment only?

MS. MURDOCH: Pre-employment only.

MR. LINEWEBER: The CDL situation/issue is a non-issue here?

MS. MURDOCH: Actually, they require anyone who is applying for a job to go through this testing program. If the person is going to be a CDL holder and applies for a safety sensitive position, they undergo a DOT pre-employment test separately. Separate collection, separate everything.

CORPORATE EXPERIENCE WITH NIDT

DR. WALSH (Walsh Group): I've been asked to talk about the corporate experience, and what I thought I would do is try to give you a case history of a company we have been working with, converting them from lab based drug testing over to the on-site program. We are only about halfway there, but to give you a sense of some of the trials and tribulations and some of the benefits of an on-site program. A lot of what I'm going to say Julie has already said. Our program is on a much smaller scale, but it will give you a sense of what the reality is.

The company that we are working with is a non-regulated organization. We don't have to follow the HHS standards. It's a manufacturing company. It has 11 plants or office sites in eight different states, about 1,400 employees, Union, non-Union, salaried and hourly folks. Six of the sites are currently set up for lab based testing. Five sites are using non-instrumented drug tests and our goal is to convert all sites over.

The company policy calls for the full range of testing policy options, pre-employment, post-accident, random and return to duty. We are actually manipulating the random testing rights here as part of a study to look at the impact of manipulating the random testing rates on a variety of dependent variables and hopefully within the next year or so, we will have some data we can talk to you about that.

What we tried to do was set up what we conceptualize as a forensic on-site drug testing program. We also selected the TesTcup 5 of the devices available. From a corporate perspective, they wanted an FDA device with minimal specimen handling and no reagent mixing and so on.

All presumptive positives are sent to a SAMHSA certified laboratory and they are re-screened by immunoassay and then confirmed by GC/MS. In essence, a confirmed positive has been analyzed no less than three times before it is reported out by the MRO. All lab positives are reviewed by the MRO. We have a quality assurance program which I'll talk about in a little more detail.

We have developed a handbook, an SOP, and we have tried to develop a consistency in the procedures and programs across all sites. As Julie mentioned, some of the states have created some problems unfortunately because this is not a Federal agency. The corporation does have to comply. We have one site in Ohio which requires a split specimen. We have a site in Connecticut which prohibits random testing. These are things that from a corporate perspective, the H.R. people have to be careful about and make sure they are complying with all state laws.

In terms of training the testing personnel, we use a training video. We have the SOP handbook for reference. We also have the 800 number and materials. We have practicum where the specimens are actually run.

We also are now beginning to train in adulteration testing. We have been doing a little adulteration study on the side, and I'll show you some data on that a little bit later as well. We have a bi-monthly quality assurance program.

Testing procedures. Basically, the preparation of the collection site is no different than for a standard HHS collection. Identity check, identification required, basically a secure collection site. The specimen is collected in the TesTcup. The TesTcup is opened in front of the employee. In fact, in most cases we hand it to the employee and the employee opens the package. The lots and expiration dates have been checked prior to giving the pouch to the employee. The employee provides a specimen, returns it to the collector with the lid off. The collector does integrity checks, looks at color, checks temperature, and in the one site where we are doing the AdultaCheck IV, the strip is introduced into the specimen. If the specimen is normal, the lid is put on. The test is run. While the test is running, the chain of custody form is filled out. When the test is completed, the test valid lines appear. The cup is sealed. The chain of custody seal is put on it. The donor signs the seal and then is dismissed, and then the results are read after the donor leaves the site. If the test is negative, it's discarded. If the test is positive, it is ready for shipping. I like Julie's idea of eliminating some of the paperwork. We have played with the idea of an abbreviated chain of custody, but at this point, this is the procedure we have been using.

We have a little QA/QC program for the operators. Basically, it's an external evaluation. Challenges are sent bi-monthly. Usually, it's two specimens, a combination of positives and negatives. Sometimes a single specimen may have two analytes positive, maybe negative as well.

The individual at the site runs these two specimens. If it's negative, checks off negative. If it's positive, they check off what it is positive for, faxes it back to us. If both specimens were correctly analyzed, they get a nice E-Mail from us saying good job. If they have made a mistake, they get a rapid phone call to find out what is going on. We try to go through it in great detail.

This is not real sophisticated but it gives us a nice, warm fuzzy feeling, at least every couple of months that the analysts are doing the job properly. So far, we have had no errors among the collectors.

There is an oversight role in here, and I think you have heard it from Bob and Leo Kadehjian and also from Julie, in terms of operating a large multi-site program. Training is critical. Program and site evaluation is also important to make sure that at least initially the sites are set up properly. We also monitor lab and MRO reports and have a system so that if we get an on-site positive and the specimen fails to re-screen or fails to confirm positive at the laboratory, we are usually on the phone back to the lab. We often will redirect analysis to GC/MS, if it fails the screen, to try and get a good sense of why an on-site positive is not confirming.

We also run the QA program for the five sites that are currently doing on-site testing. We actually do the random selection. We have a random selection process where weekly, we contact the H.R. director at the plants, these are the people that need to be tested this week, so we know who is going to be tested under the random program, and we can watch for the results coming back. We also coordinate with Corporate H.R. because ultimately Corporate H.R. is the one who is taking the adverse actions. In this particular corporation, a confirmed positive drug test is grounds for termination. The CEO of this corporation is a friend who I have worked with for more than ten years. He feels very strongly that the policy of one bite of the apple, of a positive test being a referral for EAP is a mixed message that says it's okay to use drugs until you get caught, and then once you get caught, then you need to clean up your act. This particular company has a very aggressive EAP and education outreach that says if you have a problem, come forward, we will help you. They will pay for all the costs of EAP and treatment. The down side is if we catch you, you are going to be gone. He sends a very strong message.

Back in January after one of the DTAB meetings, Bob Stephenson and Donna Bush were expressing concern over the increasing rates of adulterated specimens being received at the lab and they asked us if we would do a little study to try to determine the prevalence of adulterated specimens coming through the system. We began at one site where we had a registered nurse who was the collector, and began to use the AdultaCheck system.

This is kind of a summary of the first part of the year. What it says here is that out of the 219 non-instrumented drug tests that were done between November and April, we had five specimens that were identified as being adulterated. The little shadow across here, these are the four AdultaCheck tests. This first specimen had a low pH. It was sent to the lab, was positive. It was positive on-site even though the pH was low for THC, confirmed positive for THC. This specimen was low on creatinine. It was negative on-site and was sent to the lab. It was negative. This one had a low creatinine and an abnormal positive for nitrite. It was a dark red thick urine was the comment that was made. It tested negative in the TesTcup. We don't know why she didn't send it to the lab. What she did was she collected another specimen a couple of days later which also was adulterated, low creatinine, abnormal positive on the nitrite. It was negative on-site. It was sent to the lab and pyridine was detected by the lab in their adulteration panel.

Similarly here, 108 and 139 are two specimens from the same donor collected a couple of days apart. Here we had an abnormal positive for nitrite. It was negative on-site. For some reason, she didn't sent it to the lab. I know a lot of you have probably seen that bumper sticker that says stuff happens or something like that. I don't think it matters whether you are running a lab based program or an on-site based program. Stuff happens.

When you have an oversight role and coordination, at least you begin to pick these things up and you can call up the collector and say why didn't you send this specimen. I was going to get another one and then I couldn't find him that day so I couldn't get him back until the next day and so on.

This is about a 2.2% adulteration rate. The other reason we focused on this particular site was we had heard back through the H.R. director and the nurse that rumors were rampant at this particular plant that everybody was beating the drug test and that people were buying things and helping them to do that. Word has also gotten around pretty fast that we are now running an adulteration check out there and this is really dropped down to almost zero now.

This is pretty much second quarter 1999, on-site test results. We had 267 tests conducted. There were 10 presumptive positives, 3.75% rate, all presumptives were sent to the lab. Five confirmed positive, one arrived at the lab with less than 30 mL, so it was not tested. Four failed to re-screen. Three of them turned out to be morphines. Stuff happens even with the oversight role, and that is the overseer failed to take notice of the memo we got from the lab telling us they were shifting over to the 2,000 screen, and we were still testing on-site with the 300 nanogram cups. Roche now has a morphine 2,000. Needless to say, that was corrected within a week, but on the fourth specimen, it was THC positive on-site. We redirected it. It failed to re-screen positive. We directed the lab to take it to GC/MS and it yielded 12 nanograms of THC.

I think as we are beginning to get control over the system, we really believe we now have a forensic program which is operational, which is able to detect problems when they do occur. We are able to respond quickly for re-training issues. There will always be new collectors coming onto the scene that will have to be re- trained. If you have enough checks and balances in the system with training and coordination and evaluation, I think you can get a good sense of comfort about the forensic nature of the program.

From a corporate perspective, the H.R. director is delighted. The principal benefits are the rapid turn around time, a similar situation where a number of recruits in a tight market are being lost in the interval between when you finally interview somebody and find somebody you want to hire, you send them off for the drug test, the two to four days it may take to get the test result, the guy takes another job. Several of the plant managers who have shifted over from the lab based testing to the on-site testing have expressed extremely positive remarks back to corporate headquarters. From a corporate perspective, and I have not seen the statistics, but they believe they are saving a considerable amount of money.

The actual laboratory costs of the on-site device and the re-screening and confirmation is about the same as what it would cost with the lab based system, but where the balance gets tipped is in terms of the administrative cost savings over having the collection portion of the chain of custody there and then waiting for the test result and then reconciling, waiting for the MRO to make contact with the employee, and all of that has been significantly diminished so they are only dealing with positives. They believe that has saved them an enormous amount of money administratively.

MS. MURDOCH: Mike, you used NIDT for random testing?

DR. WALSH: Yes.

MS. MURDOCH: What does the company do with the individual if they screen presumptive positive?

DR. WALSH: This company, almost all of the plants I would consider to be hazardous jobs. Heavy chemical exposure, heavy materials. They have made a decision that on a presumptive positive, that person is put either on administrative leave or is reassigned to a non-safety sensitive position until the confirmation comes back.

MR. PORTER: You mentioned that the SAMHSA certified lab was testing opiates for 2,000. What percentage of the laboratories now are testing at 2,000 versus the old standard, which is 300? Is that a fair question?

DR. WALSH: In order to be compliant with the Federal regulations, they should be testing at 2,000.

MR. PORTER: All of the SAMHSA certified labs are testing at 2,000?

DR. WILLETTE: Many if not all probably still offer the 300. Is there any certified lab here that doesn't offer 300? They have to do 2,000 for Federally regulated samples.

DR. KADEHJIAN: Today we heard from Julie about the testing being performed in the presence of the donor. You made a point to say that the specimen is collected, sealed, and then the test run immediately thereafter, but not in the presence of the donor. Could you comment on why that is and any benefit?

DR. WALSH: Actually, the test is run in front of the donor and he is filling out the chain of custody. Just the results are not read in front of the donor. It was a corporate decision made over concern about confrontation and so on. The decision was made to go through what Julie would argue is an unnecessary step of filling out the complete chain of custody and sealing it and preparing it to be shipped to the lab. We go through that process simply so that the employee can be dismissed because everything is done. They read the results. If it's negative, they discard it and if it's positive, it's ready to be put in the pouch and sent. I think these are the kinds of nuances about the on-site programs that only experience -- going back to the Bubba argument, it seemed plausible, but in practical reality, it doesn't seem to happen. In fact, in almost every case where we have had a positive on-site and a person where they have been an employee and they have been reassigned into a safety sensitive position, I would say almost all of them have gone to their locker and packed their stuff, you caught me, and I know the rules. We have not had any arguments or any lawsuits or any confrontations.

DR. BUSH: I feel compelled to do this because we were just talking about how HHS raised the cutoffs from 300 to 2,000. The rest of the story is in the laboratory, we do confirm also for 6-AM when the morphine confirms at 2,000 or greater. That is the rest of the story. I'm so sensitized to this because everybody talks about how we raised the opiate cutoffs, but we also required additional testing for 6-AM.

DR. WILLETTE: Our next speaker is Wanda Boone who is the National Director of Quality Assurance and Regulatory Affairs at LabCorp. She has been at the Research Triangle Park location for a number of years. She is going to talk to us about a disseminated on-site testing program at collection sites.

DRUG TESTING PERFORMED AT COLLECTION SITES

MS. BOONE: Yesterday, we talked about CLIA regulations, and you don't have to read this whole slide. These slides are multi-purpose. Please bear with me. In terms of CLIA, of course, Federal CLIA exempts drug testing, but there are some states that require stricter requirements than Federal CLIA, and I think that was alluded to a little earlier. When we are talking about setting up drug testing in a national program, we do have to look at those specific state requirements. Also, Dave Evans mentioned state laws and in talking about state laws, he did mention the voluntary drug free workplace laws and some of the other requirements, but one of the things that I also want to mention is that some states require training by the manufacturer. Some states require training by a certified laboratory. There is one state that requires that the donor views the test results. The others don't have that criteria. Then there are some states where the testing can only be done by the employers.

In looking at that, I have a slide that is lovingly called regulatory soup. The reason that I call it such is because at any time, one that is creating a national program, and I'm the on-site director for the on-site program at LabCorp, we have to look at all of these various issues in order to make sure that we are legally covered. Because of this, LabCorp has provided a program of training and certification in order to meet the requirements of these various regulations, et cetera.

Let me just go over a little bit about what that program entails. We tried to make the procedures as close to Federal guidelines as possible because we offer certification on urine drug screen collections, screening test technician, breath alcohol technician, hair collections, as well as on-site testing methodologies, collections and testing. In looking at all of that and looking at what the collectors are accustomed to doing on a daily basis, we didn't want to veer far away from what they have come to know as the standard. Our training protocol, we have a train the trainer session and a collector session for all of the items that I mentioned, and adulterant testing is coming, we are looking at various products.

We have regional facility managers and support staff that do obtain their training certification as trainers. They in turn are equipped with inspection checklists on any or all of the items that I mentioned, and they are the ones that do the continuing education training portion for the entire company across the board.

The sites that have been approved and passed the inspection receive what we call a star certificate. There is a certificate for the site, a certificate for the trainer, a certificate for the collector as well, so people are really well certified. There are also three manuals that we use. I didn't bring them because they are too heavy, and I'm specifically talking about on-site collections and testing.

The on-site manuals are three different types. One for the end user, and I'll explain the differences in a moment. One for the LabCorp Patient Service Center, and one for the LabCorp Patient Service Center that may be in a CLIA type state. The only difference is just some verbiage in terms of what is required by that particular state and also there are familiar terms that we use in drug testing that may not be familiar to the end user, and those have been massaged a little bit to help the end user out in understanding that verbiage. We have a monthly continuing education teleconference where on that teleconference, we talk about changes to state regs and Federal regs, problems that we are seeing in the facilities and corrective action for those programs. I'm able to announce upcoming training programs and talk about trends in terms of what we are seeing in the on-site testing arena, and there is a question and answer period. We also have an 800 number that has a small refresher course on it where people can call in if they just simply have forgotten where their manual is or for some other reason they want to have a quick check on the process they are using for on-site testing. They can call that 800 number and hear a very short description of what they should be doing in terms of the collection. Also we have a help desk that is available 24 hours a day, seven days a week, so that if someone is in real trouble at the collection site, they can get an answer to that question as well. That does seem like overkill. I know that people are probably thinking, my goodness, why do you do all that. I'm sure you know that turnover is really high in patient service centers, and we just want to make sure there is enough information available to the collector so that whatever they are collecting, whatever medium, whatever process they are using, that they will have the tools that are necessary and the information at their disposal so that we won't have problems when specimens come to the laboratory or clients will continue to be satisfied with our services.

In terms of the collection site, you will hear some familiar items here, when we talk about security preparation, privacy, and yes, observed collections. That is one of the things that really surprised me, and this, in terms of direct observation and the shy bladder procedure, is really client pushed. It isn't that LabCorp imposed its regulatory opinions on the client, but the clients came back, even for on-site drug testing, to say to us, we want observed collections as we see with the Federal sector, and we want the shy bladder procedure followed as well. That is one of the things that we have addressed for them.

In terms of good laboratory practices, some of the things we want to talk about are security standard procedures, proper donor identification, chain of custody, temporary storage areas and transport to the laboratory. I will go through these in somewhat detail, and all of that, of course, is to ensure that we have specimen integrity.

In terms of specimen security, and again, this is familiar hopefully to many of you, sites that are dedicated solely to urine collections must be secured at all times, and this is what we expect to see and this is what is on our checklist as well, in terms of what we expect to see at the collection site. When public restrooms are used, the facility must be posted against access. The specimen must remain under the control of the collector. The donor must be allowed to provide their specimen in privacy. The donor does not have an opportunity to see their result read under our system. Unlike some other agencies where there may be control, you have control of the collectors and control of the clients or control of the results, our collectors and the third party collectors that we contract with absolutely their cry is we want the donor out of there before we read the test, and so as I mentioned as well, there is only one state, and I'm not here for regulatory interpretation, so I won't give that state but perhaps you can get it from David Evans, there is one state where it is required that the donor view the results, but that is the only state that has that requirement. The donor must be allowed to provide their specimen in privacy, and then I'll talk about special circumstances. Just as an example of what the posting may look like if you are doing a collection in a public place, of course, you want to say what is happening, the date, the start time, end time, and of course, they do not enter because that's the purpose of that posting.

One of the other things that we want to make sure of is that procedures are established at the site. Our training materials and all of the information that we have, of course, goes through this as well. Ensure that client requirements or state regulatory requirements are documented. Ensure that the facility and restrooms are properly prepared. I will go into a little bit of detail about what "prepared" means. Ensure that appropriate site supplies are available; thermometer, coloring agent. Ensure that the name of the employer representative, a very critical person, that name is available. In terms of the thermometer, in the event that there is a temperature excursion and that excursion differs by more than plus or minus one degree under DOT, 1.8 degrees under HHS, then you give the donor an opportunity to have his or her body temperature taken, so the thermometer is not to put into the specimen. The thermometer is to measure the donor's oral temperature. We have standardized on the DOT plus or minus one degree because of what our collectors are most familiar with.

Also in terms of specimen integrity, you can use a coloring agent, it doesn't necessarily have to be blue, but you want to avoid yellow and orange for the obvious reasons. In any case, you want to make sure that the toilet bowl remains blue. The water in the toilet bowl remains blue, tape faucets and shut off water supplies, hot and cold. Again, the reason that we go through this much detail is because this is what the collectors are accustomed to as they are collecting Federal specimens. We want to make sure that they don't deviate from that procedure, back up from that procedure, and then make mistakes with those specimens as well.

After the donor has washed their hands prior to the collection, then you want to make sure that they don't have access to anything that they might use in order to adulterate the specimen. Then you want to check the specimen temperature within four minutes of receiving it from the donor. Any time there is a suspect specimen, and I mentioned the temperature issue, but any time there is a suspect specimen, whether the temperature is out of range or there are foreign objects, unusual odor, blue dye, then you want to document the observations on the chain of custody form, concur with the supervisor and contact the ER, which is employer representative, not emergency room, of course. I mentioned what we do if the temperature is out of range. That would be a situation where direct observation might come into play, but in any case, regardless of whether it's an abnormal on an adulteration strip or there is some other issue with the specimen, those specimens are packaged and forwarded to the laboratory for further testing. Again, even in on-site testing, direct observation was client driven. I mentioned what we do under these situations. Direct observation and shy bladder requirements have become what the employees that we have seen see as the standard, and even though it's a rapid turnaround situation, in these abnormal situations, they want to see these types of procedures followed.

In terms of shy bladder, again, there is a little difference between DOT and HHS on this as well. HHS remains eight ounces of fluid every 30 minutes, et cetera, but of course, the DOT procedure changed a couple of years ago, and it's 40 ounces of fluid over a three hour period. Many of our clients do want the donors to stay at the site and drink fluids for those three hours, even though again, we are talking about a rapid test. That is what we use as the shy bladder procedure.

In terms of the specific collection procedure, and you have heard this several times over the past couple of hours, donor identification is important. Chain of custody is important. Donor preparation is important. Specimen integrity, of course, is important. I will just talk about those aspects for just a moment. In terms of identification of the donor, we want to make sure the donor is identified through a picture I.D. or through the employer representative, and the use of non- photo I.D.'s or anything like that again is unacceptable. We are doing this again in order to make sure that our collectors are doing things consistently as far as we can help them with in terms of the collections they do for DOT, HHS, NRC and in this type of arena as well.

We do use the chain of custody form. I'll show you one in just a moment. On the chain of custody form, the donor does not sign the chain of custody form just for consistency sake. The collector must enter the date for each transfer. The collector must sign and print his or her name for each transfer. The collector must enter the purpose for the chain of custody entry for each change, and the collector must sign and date certification statements. The donor also has a certification statement to sign in using the chain of custody form that we do use.

The chain of custody form that we use is a five part form, and I won't be able to do this very well. Those are examples of positive and negative forms after they have been completed. In terms of the temporary storage area, we have to ensure, or under our program, we ensure that the area remains locked to prevent unauthorized access to specimens and show limited access and then if there is a key, exercise key control.

In terms of transport to the laboratory, again, familiar place, the chain of custody in the designated area of the shipment bag, place specimens in the designated area, ensure that the absorbent sheet is there, seal the bag with a shipping seal, store specimens in a secure area until pick-up by the courier.

I'm walking over here because I do have a packaged up kit that I'll pass around as well so that you can just -- and I'll just start that going around as well in case anybody wants to take a look at, you know, the completed kit with the test cup in it. In terms of the collection device, we do use TestCup and I'm not going to go into a lot of detail about it because you've heard about it from Sal yesterday. But in terms of looking at the collection device, we're talking about the collection container capability to secure FDA clearance and impact on the specimen. If this was not made clear, well, you see RADAR on all of these slides and RADAR stands for rapid assessment of drug and alcohol results. I didn't mention that earlier, but TestCup is a product, one of the products of this program, specimen collection and testing in the same unit. No contact with specimen, simple, and easy to interpret results, screening tests completed in less than 10 minutes. Change of custody and GC/MS confirmation. Again, you've heard about that many times so I'm not going to go over that. I have a slide on the product description but you've heard about the product description, so I'll just bounce over here to one of the slides on the principles of the test just for memory's sake.

Just to show that one of the things we do is to wait for the test valid ban to appear. As that packaged kit is going around, you'll see the test valid bans on the test cup and actually that is not a positive result that is being forwarded to the laboratory, it is a whited-out positive. But, in any case, you'll see the test valid bans and what they look like. And, if the test were negative, then the negative test cup result would yield a blue line, blue band as you've heard, and if positive, it would yield the colorless white symbol. You will see that with the test cup that's coming around as well.

In terms of the specific procedure for collection, again, this is kind of a summary, because I've been through many of these already, in terms of going through the general sides, but separate and apart from the training program and certification program that I mentioned, the package insert and test procedure, transfer materials, chain of custody, temperature strips which are now on the test cup are a part of what the end user gets.

The lot and expiration numbers are noted in the remarks section of the chain of custody forms and you'll see that on the forms that have been completed that are going around. QC, in states where CLIA is an issue, then what we must do in those states is to ensure that there is a daily negative and positive QC that's done on each day of testing. That's for those particular states. Otherwise, we test a positive and negative with each lot of 25 TesTcups. We do have a double blind QA program that includes monthly submissions to the patient service centers and we're developing a program for negatives to be sent in as well. But in any case, you want to make sure that the collectors have all their supplies, where they have the chain of custody form, the test cup wrapped in the white plastic and there's one in there, if you can start that going around, they note the expiration date and lot number on the chain of custody form. If the temperature strip is not on the TesTcup and as I've said, the temperature strip is affixed to each of the TesTcups now, we'd affix a temperature strip and make sure that there is the transport bag and overnight mailers and the use of medical gloves is suggested. In preparation again, we want to follow all these procedures, take water sources, add coloring agent, remove cleaning agents from the restroom, explain the process to the donor, and we do have sheets that are used for the donors both in Spanish and in English. Ask donor to remove outer garments, provide locked storage for their belongings, obtain the donor's photo I.D., the donor's social security number, and instruct the donor to wash their hands prior to the collection.

In terms of manipulating, this is not manipulating the TesTcup yet, but in terms of the chain of custody form, I mentioned briefly the steps, so I'm not going to go over that. You can see the steps as the form is going around. But the donor is to void at least 30 milliliters of urine directly into the test cup; 60 milliliters if a split specimen is collected. If a split specimen is collected, then 30 milliliters of that 60 milliliter void is poured into a transfer vial. The chain of custody form is completed as I briefly went through it and the temperatures read within four minutes of collection and documented on the form as well.

They want to examine the TesTcup specimen for signs of adulteration as I talked about earlier. Then they're going to secure the TesTcup lid, turn it to the test position, and tilt, and you had that demonstration yesterday afternoon so I won't go over that. But then you wait for the test valid bands to appear. When the test valid bands appear, and you're ready to read the test, then you turn the lid to the lock position, apply the security labels, have the donor initial the collector dates, the seal, and then that donor can leave the site. So place specimen label A over the test cup lid, have the donor initial the seal, the donor may leave after verifying that the specimen I.D. numbers match. So they have an opportunity to look at the number that's now on the test cup and the number that's on the chain of custody form which they signed attesting to the fact that it was their specimen. So they have an opportunity to do that.

They do get a copy of the chain of custody form before they leave the site. This is just another slide that really just says what I just went through, turning it to the test position, inverting it, waiting for the test valid bands. There's a plastic strip that reveals the result. The collector records the result on the chain of custody.

As I mentioned there are blocks on the chain of custody form where they check negative or positive for negative result -- blue, positive result -- colorless.

In terms of the chain of custody block, if the specimen is being transferred to the laboratory, then the collector completes the chain of custody form as they are accustomed to doing and they discard the negative specimen and you can just see an example of how that chain of custody is filled out for the transfer to the laboratory. Well, how much more could there be? Not much. Place of the chain of custody in the designed area, as I mentioned, and you're seeing a sample of that going around. Just one more slide. In our opinion, and my opinion, the key elements of a successful on-site program, we do have a training video, but the training certification is really the meat, I feel, of the program in terms of keeping people updated on what's happening. A lot of the continuing education that we do may not be directly associated with on-site testing or breath alcohol or screening test technician, it might have something to do with an article that was read that's very interesting. But the reason for that is, and we do keep that as a component as we're going through all of the other information so that it will raise the heights, if you will, the thought of the collectors so that they can think up here about what they're doing instead of seeing themselves as being low on the totem pole.

We have found that when we've given that much attention to the collectors in terms of the monthly teleconferences of the 800 numbers and all of that, that they do a fantastic job for us. This program has been in place for about 3 years now and we haven't had any problems. We resolve the problems that we do have in terms of concerns about collections when we are on those teleconferences or in the phone calls that come in. That communication has been extremely valuable. We feel that chain of custody is very important not only from documenting information on that form in terms of the result, but because we want to have consistency in what we are doing throughout the collection processes for Federal specimens or non-Federal specimens. The collection and testing may be performed at patient service centers or employer sites. I think that in terms of manufacturer training, that is certainly acceptable but the laboratory or someone should be brought in to talk about forensics as well and that certified laboratories be used for confirmatory testing.

MS. MURDOCH: I have two questions about your chain of custody form and your procedures. I noticed in your procedures it said that they are supposed to fill out steps 1 and 7. But step 7 is the certification of the individual that the specimen has been sealed in his presence. How can you do that at first?

MS. BOONE: The reason that we do it at first and actually, that's a statement that needs to be changed. I'll put that on the record. That statement is used for our routine urine drug screen, so that statement needs to be changed for on-site methodologies. They will authorize giving their urine for that process.

MS. MURDOCH: The second question is you said that the results are recorded but I don't see any place that they are. I see some place where it says test requested by employer, and then under that it says negative, positive, or profile one. But nowhere does it say, I, the collector, got this result. Nowhere does the collector -- it doesn't seem to reflect the result. It reflects the test requested by the employer.

MS. BOONE: Probably, what you need to look at is the completed chain of custody forms that are going around, because it does have the check by the collector and should include the collector's initials. You're just looking at a blank form.

MS. MURDOCH: What does test requested by employer mean?

MS. BOONE: There are times when the employer may request other tests other than the TestCup and so those can be identified on that form as well. It's a multi-purpose form.

MS. MURDOCH: It's very confusing.

MS. BOONE: That's why we have extensive training.

DR. KADEHJIAN: You indicated earlier that you had a large number of collection sites. What was that number?

MS. BOONE: We have about 800 that do forensic drug testing collections.

DR. KADEHJIAN: And do you have data on positivity rates, confirmation rates, any of that?

MS. BOONE: Yes. I don't have it with me, but yes.

DR. KADEHJIAN: A rough idea?

MS. BOONE: I'll get back to you on it. I have your E-mail address.

DR. KADEHJIAN: Thank you.

DR. WILLETTE: Our next presentation is by Ted Shults. He is going to talk about medical review officer issues in regard to on-site testing.

MRO EXPERIENCE WITH ON-SITE DRUG TESTING

MR. SHULTS: I want to thank Bob Stephenson and Donna Bush and the rest of the Office of Workplace Drug Testing Programs for the invitation to come here to this meeting and present the prospective and experiences that medical review officers have had in this area, and as most of you know, one of the hats that I wear is the chairman of the American Association of Medical Review Officers, which is one of the two bodies that provide certification for physicians in this area. Certification that has in the last couple of years, been recognized by state legislators and more importantly, we have a couple of states that require it now, but more importantly by the marketplace. Without there being a mandatory requirement for this, I would say a great deal of drug testing is done today is done by a certified medical review officer.

I've had a number of different perspectives on testing for a while. My own training is originally in toxicology and I did develop methods for the detection of drugs in blood and urine. My donor, however, happened to have been a thoroughbred racehorse. I worked in the laboratory that did racetrack testing and it was some very solid lessons that I learned in all of that. But for the last ten years or so, I've worked as corporate counsel for one of the large testing laboratories and I've handled countless drug cases, from Federal cases down to Worker Comp cases to unemployment security hearings. I've had some degree of exposure to all of this, but I always learn a lot.

I learned a lot in the last two days and I think that many of you have learned a lot as well. One of the perspectives is that this is a rapidly evolving area. They're trying to like characterize the Internet. We are in an explosion of technology and an explosion of interest in this. It's not just on-site urine and oral fluids but other technologies as well which the DTAB group is trying to assess. I apologize for the presentation. I feel a little bit like NASA must of felt a month or so ago. You know, the issue that they had that half of the researchers were working in pounds and inches and the other half were working in kilos and centimeters. Well, my computer is working in pounds and inches and that gizmo over there is on a different framework. So I'm a little bit like, well, we're launched here, but I hope I don't get going into orbit too much. You're only seeing about three quarters of the slide. But, in any event, that may be enough to get it across.

There is one thing I wanted to do is share with you sort of a visual image of all of this. I also like to do a little historical background and find an appropriate kind of quote, you know just to sort of set the tone for it. And after Bob had asked me to do this, I heard the quote. It is actually a presidential quote, but I think it encapsulates what we have been doing here, what we're all about, and what this process is all about.

The quote is -- it really reflects a type of philosophy and that is, "To centralize ideas and to de- centralize execution." That is what a lot of what on-site testing is all about and it certainly is what we're about here. Centralizing what the concepts are and the basic principles from a process that will go down to a much lower level. Now, unfortunately, the president that's quoted, or who held that philosophy was Herbert Hoover. But apparently, and I was watching this review of him, he is somewhat underrated. I think there may be something similar there as well.

I think on-site testing in many regards has been something that has been underrated. In terms of the medical review officer, I think there is a overly-simplistic review of what an MRO is all about. I think they are supposed to just determine if there is a prescription and call it a negative and if there's no prescription we call it a positive. Verifying results is certainly the core piece of what medical review officers do. Keep in mind, they are where the rubber does meet the road. They interface with the donor, they interface with the laboratory, they interface with the employer, and actually, from the prospective of this type of product, they are probably able to provide, whether you call it point of care, point of collection, or on-site testing, services in the marketplace right now. Because why? Because there are a lot of MRO's.

We look at it here, a national type of level, with major well-run programs, the Post Office program, what goes on in criminal justice, these are very well-run programs. They are large programs. They have a lot of specific needs and difficulties that they have to overcome. A lot of testing isn't like that. A lot of it is very local. There are quite a few medical review officers out there. There are probably 5,000 certified doctors, but that only reflects, I think, the tip of the iceberg, because what those certified doctors are also are that they are the local expert. Other physicians go to them for their opinion and their assessment and the general growth of the medical review officer practice has been driven to a large degree by non-regulated testing. Most of the physicians that come to our training programs are not even doing DOT work. The majority of work that they are doing is non-regulated testing.

The Americans with Disabilities Act has profound implications from medical review and I will tell you if you are not using a medical review officer, you are probably not in compliance with the Americans With Disabilities Act. You're dealing with personal information, personal medical information. There are just some generic state law confidentiality issues, privacy issues. There is a movement towards looking at fitness. We're looking at the issues of performance and prescription drug abuse and return to duty issues. There is an intangible, immeasurable value to the MRO's expertise, medical expertise and common sense. They are looked to for their professional opinion in a lot of these areas.

You know the companies that don't use medical review officers, they say, yeah, but what the MRO also does is it insulates that donor from the management. That really is probably one of the benefits that companies don't really appreciate until they actually begin to use a medical review officer. When you think about this, and you think about those roles, and you think about the emergence of the type of technology we've been talking about here today, and the technology that's on the horizon, what you find is that there is an emerging role. The emerging role is the MRO as a drug testing consultant and facilitator.

I know that a lot of marketing and business strategies are based upon the displacement of laboratory-based urinalysis testing, because we're going to replace it with oral fluid. We're going to replace it with hair, we're going to replace it with ear swaps or nasal discharge -- whatever the specimen happens to be. But the reality is that all of these processes have their particular strengths and weaknesses and ultimately what you will see is that this will become part of the tools that companies and physicians use in assessing specific types of problems and solving specific types of problems.

MRO's generally look at this as an exciting opportunity in the sense that they've recognized -- I'll talk about in general terms what some of the experiences have been, that there are a number of employers that look at point of care, point of collection testing in terms of the increased turn-around for negative. We've had a number of presenters mention that element of it. It is not necessarily looked as a replacement, but often now being looked at as an option.

If you need to make that hiring decision now, this would be an option, and they're willing to pay, as is the Post Office generically is willing to pay a higher price for that. It's also used in a number of other applications, not necessarily workplace applications. You will see that this is something that will be useful in the in-house medical evaluation assessment of adolescents.

One area of growth for physicians is the family that comes to the clinic and says we're concerned about our kids, we'd like to have them assessed. Now the MRO doesn't necessarily think that this is a be-all and end-all tool for an assessment, but it is something that the family is interested in getting and they can do a clinical assessment and probably get more information doing that, but since the family is willing to get a positive or negative drug test, this is a procedure. This is something that is being used more often. Of course, there has been a good deal of history. There is hardly an emergency room in the country that doesn't have some sort of rapid assessment tool to assess clinical presentation situations.

When we talk about urine tests and we talk about that aspect of point of collection testing, in one way, it's easy. It's one way there is no resistance to the idea that the interpretative issues will be different. Why? Because if you are depending on a laboratory GC/MS confirmed result, the issues are identical to the existing process to a large degree. These don't present necessarily a different spectrum of interpretation or different areas to be considered. Now, obviously, that will be slightly different for oral fluids, there's a slightly different window of exposure. There may be some other technical variations, but urine is urine and if the positives are going to a laboratory, then the interpretative issues should be the same. So that probably is the least amount of resistance. There's also the opportunity here as many of you have mentioned to address adulteration from a different prospective. A little bit bringing it down into the field.

I just recently saw a paper from out of Oakridge, Tennessee, of a bunch of chemists who are not necessarily toxicologists, that basically, I think, the conclusion of their paper was that many of these oxidizing agents work rather rapidly, which surprised me, because I was always under the impression that in a neutral environment, unless it was very acidified urine, that these were relatively slow processes, measured in days and actually the kinetics experiment that was performed in this paper seemed to demonstrate that some of these oxidizing agents, particularly the nitrates, actually work at a much faster rate of reaction than we had anticipated. And, again, and my other hat, as a consultant, I have been a legal consultant to about half of the nuclear utilities and I have looked at their on-site testing programs and many of them had used the availability of specimen validity, have done some pH testing, have done some types of specific gravity readings, with even refractometers, with the idea that with the diluted specimens, that may be a way of managing that.

I think that all of the presenters have basically mentioned this, and I think my recollection is that nobody didn't recommend that adulteration testing can't be enhanced from this approach. Again, as we know, there is a degree of gamesmanship involved in all of this and there's more than one way to skin the cat.

I should say that about all of the things that we've talked about. I think there are a lot of experts in here. There are a lot of good opinions in this. The marketplace is another factor and there is more than one way to skin a cat and I think the marketplace will dictate to some degree as will the courts as to which ways are the best.

When we talk about general concerns with medical review officers, you can say, oh, I know what they're concerned about, they're concerned about losing the negatives. Well, that's not necessarily it, because there are some large MRO providers, but for most MRO providers, this does not make or break them. They're doing this as part of their whole medical practice, part of their occupational practice, part of their health care delivery system and negatives this way or that way are not that big a deal. They are, however, concerned about testing without confirmation. Of course, they are worried about testing without any medical review, positives or negatives. You say, well, how could that happen, and I'll explain that in a minute.

The general erosion of safeguards. I mean, to a large degree, most of the technical issues we have been talking about over the last two days have involved quality control, quality assurance issues. The erosion of ethicacy with the whole program, perhaps the increased legal responsibility for actually doing a screening test, and there's another little element that I'll talk about briefly in terms of actually a conflict of interest that some MRO's have raised. This is my visual image of all of this. I said what picture can I talk about from a perspective and this is not a good depiction of it, and it's actually from a Mercedes Benz commercial, and the rest of the advertisement is, and I don't know if this will come across, but the only word in this advertisement was "trust." I think that's an important piece.

When I first tried looking at it, I said to myself, well, gee, that probably is a perspective of a lot of medical review officers, but actually it may be the other way around. That may be the MRO when it comes to all this and I'll explain why. Trust is an important part of this product. Drug testing is an intangible product. It is a complex product. It is a product that the lay individual is unable to access. They don't want they're getting delivered until it's not delivered. Even the courts struggle with trying to assess this. We struggle with trying to assess some of the technical issues of all of this. But it has been built on a foundation of trust and confidence. That is part of what this product is all about and with that, that is the concept of trust, comes the issue of fundamental fairness and as you know, because we are talking here about a governmentally-mandated test, we are subject to the Supreme Court's requirements in this area, because essentially this is a search, and because of that historical role of the Fourth Amendment in the role of the Government, the whole concept of urine testing operates as a narrow exception to prohibition against warrantless searches and as a result of that, has been that the historical objectives of the Federal program have been to have no false positives, minimal false negatives and forensic standards of defensibility. I think we have achieved that. I think that SAMSHA and the program that was promulgated on the national certification program and its brother program or sister program at the NRC, have met those fundamental objectives.

Let me recite to you in terms of what the general standards are that emerge from this, because I do think that we have a general sense of acceptable standards, not just for Federal testing programs but for testing programs in the United States. This statement comes from a little outfit called the National Academy of Sciences, and their comment on all of this was "It cannot be overemphasized that without confirmatory testing and careful medical review, treating the results of urine screening as evidence of drug use is unacceptable and scientifically indefensible."

This comment I had read many years ago and forgotten about until I recently saw it again. You know where I saw it? I saw it in a publication published by the ACLU in terms of their general opposition to drug testing. But keep that in mind that again, trust is part of this program. So we open up a philosophical debate and, yes, I heard, I think, the assessment of the issue of what is defensible. Well, I've handled countless drug tests as an attorney, as an expert witness and as a consultant and defensibility is often times in the eyes of the beholder. I've seen practices that you would say were unacceptable, pass. I've seen practices that I thought were well done not pass judicial scrutiny.

What we have here is another problem. It's not a question of what is the standard. Should it be beyond a reasonable doubt? Should it be beyond the preponderance of evidence, because in private sector workplace drug testing, there is no standard. Those standards are irrelevant when there is no legal duty to the donor. And therein lies the big problem. Therein lies the slippery slope for drug testing.

Now it isn't drug testing that will be part of a Federal program because that would be unacceptable. It would not be legally permissible, but it is permissible for an employer to use a $5.00 drug screen, no confirmation testing and no MRO and it has been part of the marketing plans of many manufacturers who are also in it in the interest of market share and of selling stock to be able to promote that concept. What a wonderful world it would be if all these small companies would simply adopt a $5.00 drug test and therein lies the issue. I mean, I'm a corporate consultant and I believe in the corporate at-will doctrine but I also believe in the idea of fundamental fairness and that the idea that drug testing is built upon a concept of fundamental fairness. Defective drug testing, with that $5.00 drug test, you have undermined the public support. I get this type of telephone call. I get the telephone call over the AAMRO line from the mother-in-law; the mother-in-law who is a nurse, whose son-in-law has just been fired or just lost a job, and she will say, you know, I know that my son-in-law doesn't use cocaine and they just did an on-site test with him and they say he's a cocaine user. I know that's wrong. Well, we don't know that's wrong, but we're not sure. It raises that element of doubt.

MRO will tell you the common denominator with many drug users despite their ability to fold up the tent in a confrontational situation with their parole officer, it's not like that in the private world. They deny use. That's the essence. That's why we do testing. That's why we have legal defensibility. They all deny use and you can't tell a book by its cover that doesn't oftentimes fit the stereotype. This was not exactly a hypothetical situation. I talked to this mother and nurse for a while and I knew she was concerned. I knew deep down in her heart she was worried about her son-in-law. She doubted that he wasn't using this drug. But we didn't have a medical review officer, we didn't have GC/MS, so we didn't have that kind of assurance so I could guarantee you. She said, what should I do, what should I do. Well, I said there's really nothing I could tell her, and that's what's unacceptable. That's where we are. There is a doubt about it. And frankly, physicians and MRO's do not want to be associated with a defective product, and you don't want to be tarnished with that. The medical review officer provides a level of trust and confidence in these products.

It's interesting from a sales perspective. I've been involved with a couple of programs that just monitor how clinics implement point of care testing, and one of the places that I worked with for a while, there was a proactive physician involved, and he had his staff call up his clients and basically they would say this, you know, we've been doing a urine testing here for the last couple of years and we now have this FDA-approved device here that we can get you the results in a couple of hours -- spare me the details. What they found was they couldn't sell that, they basically said no, we'd like to do it the old-fashioned way. We're happy with that. Let us know what happens. However, when the MRO gets on the phone and says hi, I'm Dr. so and so and we've been doing your testing down here and I just want to let you know we have this new device, and we'll be able to get faster results in most of these cases, guess what, 30% of the clients decided that's the way they wanted to go. What does that tell you? They don't believe it. It is the personal trust and confidence, the element that made those clients decide to make their changes. Now the doctors who have been using this, and my own experience with this from a variety of different perspectives, has been what I think many of the consultants who are involved more proactively in the area of quality assurance will tell you. There has been a degree of high variability between batches. I mean, I've had doctors who have noticed without quality control that gee, we're not getting any more THC's. What happened? They stopped smoking. Was the program a success? Deterrents work. I said, buy yourself a couple of QC's or take something from your other side of the business and see if these things work, and they don't. They didn't. I've had that with more than one manufacturer, that just a batch of stuff doesn't work. And, again, it was just the observation that what happened, no more THC.

That sort of strikes me, and again, it's a moving target. I know manufacturing processes are improving. They're ongoing and I know all manufacturers have struggled a little bit with the THC issue, but again, it's something that, who's watching the hen house here? And then more importantly, let's assume we're the medical review officer for a large company and we implement this in 15 different locations and we don't have the luxury of a consultant to come in to standardize these practices. Part of what an MRO has to do is to talk to the donor and the donor will say, you know, a funny thing happened to me, I never saw my specimen, in fact, I had to urinate into a glass bottle, an old coke bottle. It's like, what the heck is going on here. At some point, the medical review officer does contact the collector to try to find out what actually happened and at some point, the practice may deviate so far from what HHS requires or what DOT requires, that's not a collection any more. I don't know what it is, but it's a canceled test. That's not a direct part or necessarily a standard or practice in this industry but MRO's do that.

How are they supposed to figure all this out if there are going to be 60 or 70 different practices or procedures with 20 or 30 different devices. Again, then we have the idea of which I think the universal support of the idea, oh well, we do need to have training, we do need to have certification. I mean all manufacturers' training programs vary a great deal and of course the certification process presents a nice piece of paper oftentimes but those processes haven't been validated either.

Of course, it's a conflict a little bit. You know, you're not going to have a program that's so rigorous that it's going to deter the sales and marketing of your devices. There is some challenges in all of that side of it. Probably more important has been the idea of losing control of your program. You know if you are a large program. And that is when you put on on-site testing, one of the things that you have to realize is that data management becomes much more of a bigger challenge and of course the collection site, not from a technical process, but from a procedural and administrative process, is the weakest control point that we have out here. I think generally MRO's will tell you the weakest part of the drug testing, and so will the consultants tell you, the weakest part is oftentimes right at the collection. This is what we were going to load onto them. We want them to now also do these additional procedures and practices.

How are they going to manage the data? We saw, you know, we had two or three presentations over the last two days that address that issue directly. Some of them are very elegant, sophisticated processes. But I think process is a part of the product. On-site testing is a whole lot more than the device. It is a whole program. It's data management, it's procedures, it's training and it's control. How are these collections supposed to work? I mean, I've got the, what everybody knows, sort of the quick flow chart here of how drug testing goes, from collection site to a centralized laboratory and then at least under the DOT program, the centralized laboratory reports this. Now the benefit of all of that is you can take multiple collection sites when you've got one point of control. Either the laboratory, or more often, the MRO, has got all the data -- positive, negatives, and it's a great statistical control point. Now what are we going to do?

How do you visualize this actually occurring? How do you visualize the increased number of devices, procedures and practices at the collection site? Much less, how the date is going to get distributed. This device goes here, the negatives go there, the positives go over here, this is an indeterminate so it will go that way. Good luck.

This adds a profound amount of inefficiency into the system. This will break the system despite all the benefits that we've talked about here. That one point is the Achilles heel in terms of national programs, just dated distribution. And again, what are the best practices and procedures. You know, we've heard a couple of different issues about do you confront, don't you confront, do you have a COS, don't you have a COS? To me it doesn't matter, to the degree that they're diverse opinions. It may very well be that the marketplace and experience will be able to define what is the ideal thing. Certainly, I thought about things in the abstract that didn't really work well when it got into the real world. So the real world experiences will be helpful in this area, but there should be, I think -- I know when one of the issues I get oftentimes is that when a clinic or a MRO or a collection site decides to go on its own to offer this type of service, they're confounded. They're confounded by the idea of do I have two bottles, do I have one bottle, do I do this first, do I do the chain of custody second, and then we're getting different sort of standards from the vendors or manufacturers of the devices.

Another issue that came up that I wasn't really thinking about but we had a discussion of this at the advanced program two weeks and I asked the MRO's at the advanced program to download, give me what their concerns were, and I had two concerns that, again, I hadn't thought about. That is, is there a conflict of interest in an MRO verifying the results that one of his own individuals used. And a related issue is is there a conflict if the MRO actually owns the device. Now the second one is kind of interesting because we have a conflict of interest rule in the DOT or SAMSHA model about not having the financial interest as being a separate checks and balance. I just want to point this out as an issue because I don't necessarily think that it is. Because one of the arguments that I made, well, doctor, don't your people already do the collection? I mean if you're going to have a conflict of interest with the collection, what's the difference if they're now doing the testing. Where is there an additional area of conflict? And in terms of the second issue, there has been, and I think there will be a maintaining of a separation between the MRO and the certification laboratory, but screening is different.

Screening is screening. The negatives, there may be an issue there, but there also may be more control over the process. So, again, I just want to raise that as something, again, I had not been really sensitive to or thinking about up until recently.

The other issue I hear from many of the physicians is about the Clinical Laboratory Improvement Act and, of course, we've heard a lot about that in the last two days. We know that right now at the Federal level, it is not directly applicable to this process and that the states are beginning to relax this concern and this is kind of deja vu in some regards. I mean, back in the early days of being a laboratory counsel, I used to hear from the state saying drug testing, even the laboratory drug testing was subject to the Clinical Laboratory Improvement Act. In any event, there has been somewhat of a relaxation, but keep in mind, if you're talking about health care professionals, they're used to this. They know what this is all about. And again, it may very well be, even though the states may relax it, it is not a strong barrier to implementation of this under CLIA, under future philosophy or Federal concern about how these things are being done.

Again, we've seen a very innovative if not elegant device which sort of makes this a hybrid process now that actually does the reading for you and does the result and I was joking to Dr. Lappe and said maybe you just built a little speaker that turns it back into an analog thing that says "you're fired," but Murray will have competition.

There's something interesting and desirable even though we have good evidence to say that the readers do pretty good and if they're trained, there's no difference between readers, there's something attractive about a device doing it for you. There's something attractive about a digital.

I know three or four of you were out there sort of trying to knock it off or develop a better mousetrap, but the issues that come up is that who is going to be monitoring these. This is not -- my sense of it was it is a lot more complicated than it sounds, and of course, there will be other software, but who's going to be validating and monitoring that. Now, obviously, the FDA is looking at that, but, again, as we see more of these things come out, they may not be as sophisticated as what we saw yesterday.

When you get all down to it, you put all of this stuff together, all of these concerns, we've also seen a trend in the courts to look a little bit differently at the relationship between service providers and donors in this area. You know, despite the fact that the employment at will doctrine is on pretty solid ground, there has been an increased scope of liability in a lot of jurisdictions concerning the relationship between a service provider, a medical review officer, a laboratory and a collector toward the donor.

You know, when I was chief counsel at the big lab, I used to stay up at night saying what would ever happen if the donors could come after me, you know? And I knew, well, they can't. They can't because I don't have a direct legal relationship or a legal duty to them. I have a duty certainly and a lot of contractual obligations to clients who are my business, but these donors are not my problem. But we have seen -- and I have a whole separate presentation about how that concept has been eroded -- that courts in many jurisdictions are holding collectors liable, they're holding MRO's liable directly to the donor. There is a concern here, am I increasing my risk exposure as being a provider of these services. Of course, what drives this is not our good intentions, but oftentimes the efficacy of these programs in terms of the economics. I'm a proponent of this kind of concept.

There are a couple of different models. I think what I heard that was interesting was the ability to sort of custom make your program as the parole and probation programs have, the flexibility in all of this, the idea that it is a menu item, if you need this testing fast, you can get it, and another just economic issue that MRO's talk about that you may not appreciate is that they warehouse a lot of stuff. Laboratories send them cups and chain of custody forms, and nobody steals these. You do have to have controls over this product. This product walks out. Why? Because the staff wants to test their kids. They want to test each other. It's a party favor. The clinic that is involved with this had better keep this stuff under lock and key, not just for the forensic issues, but just because these things tend to walk out of your clinic. There is a significant amount of product shrinkage, and most manufacturers won't just sell these things the way they sell urine cups. Of course, the other collateral issue in all this, and it goes back to the black eye issue, the issue of what's the impact, the negative impact of drug testing, who is marketing all of these things. We have a lot. We have manufacturers. We have laboratories. We have MRO's. We have TPA collectors and even retail outlets. You see now going into K-Mart to be able to buy these devices. Just take all the things we have talked about and wonder about what the application, what the efficacy of that is and whether or not that will come back to be a problem as well. I guess that is who is the target markets for these types of processes. The drug using population is a market. You can buy a lot of these devices in tattoo shops in Florida, a great outlet there. If you need a drug test device, go in there and you can get pierced while you are there.

I think what we are looking at and I think what we are all struggling with is what are the minimally effective standards that we need to have in here. I will make an economic argument for you as well. If you don't get behind the idea of standards, if you don't get behind the idea of maintaining the quality and the trust of the product, you will never make any money in this business as an industry. Why? Because you will always be competing against that $5.00 product. You will always be competing it. There is an economic interest in actually what the Federal Government is doing with this program in terms of maintaining standards. Otherwise, it is always the lowest common denominator. You are always competing against all that.

I will give slides to Donna and to Bob and they can make it available for you. It may even go up on their Web site. Basically what we are looking for is a model. The problem, and I don't want to minimize this concept, which we all know, the Coast Guard is competing with UPS for package handlers, and we have an ongoing problem with drugs in this country. It's a problem that is not going to go away quickly. There is a lot of efficacy and a lot of potential for testing. It is going to have to be maintained, that element of trust and reliability is going to have to be there for the public to continue to accept this, because there are tradeoffs, there are privacy tradeoffs. There are liberty tradeoffs. There are economic costs of testing in all this. I will tell you there is a lot of frustration out there, frustration with our inability to get a handle on all this. I don't want that to sort of flood over, like the eastern part of our state, all of that negativity just sort of washing over drug testing before it really has a chance to get going. Ultimately, who is going to do this and when are we going to do this? I think we all have to do this. I think we have to get the buy in sort of together, in terms of common interests and common goals.

I know on the one hand you buy into this and on the other hand, the issue is in marketing, the lowest cost test. The economics are an important point to this and no one wants to engineer this. You have heard a couple of ideas here that you could engineer yourself right out of the market. You could have a perfect product that nobody is going to buy. The issue to some degree is what should we be doing, what should we be looking at in terms of maintaining the trust and confidence that we have built with these programs. I think what it comes down to is really looking at in terms of again a voluntary type of approach, and what AAMRO is working on with others is the idea of developing requirements of ethical workplace drug testing that apply across the board. Again, you can go with this, you can exempt yourself from it. No requirement to do that, but just that it does have these fundamental elements, these intangible elements.

What that is going to require, of course, the devil is always in the detail, is some more drilling down in terms of what are the most cost effective QC's. Not just because we are selling urine specimens to QC's, what makes sense, and that is why in some degree why this work is going to be done by DTAB. They have the experts' experts on that list who are going to be looking at what are the most effective approaches. What should we be doing at the testing facilities. What are the appropriate training standards. Who should the providers be in this. Should all the negatives go to MRO's or if not, how are you going to control that you have not lost control of your program by not even knowing who has been tested. Ultimately, you are going to need to have employers to buy into this. I think the development of those standards, which not just apply to this, this is not merely something that is going to apply to on-site urine, but I think applicable both to other technologies as well, that those fundamental principles, that we all can agree on, are going to be abided on. Of course, if you don't buy into it, that's fine. You will be sold against. Your competitors will say, you know, they are just not following the ethical standards. I think just like with MRO's, and I never pushed for the idea of making that a requirement, the marketplace did, and I think the same thing is going to happen in this area as well.

I want to thank you for your attention and the opportunity to share this perspective with you.

QUESTION AND ANSWER SESSION

MR. FORTUNA: I represent a collection site, so I'd like to make a couple of points about the collection site. We provide some options that I think need to be addressed, and one of them is for our clients, we need to provide trust. When we say their pre-employee screened negative, they have screened negative. When we come in and we do a collection for an employee, either whether we do a full DOT or using one of these on-site kits, I as a collection have some options. I can provide information to the employee -- employer about adulteration, what they can do if I suspect or detect an adulteration or temperature range problem. How to handle positives. Some of my clients do not want to use MRO's. Some of my clients don't want to use a laboratory on these on- site kits. What do we do in that situation? I then say to the donor, you tested positive. You have every right to pay for a test at a certified laboratory and have an MRO done. I give those options to the donor, I give it to the client.

We also handle the questionable ones. If I have an on-site test and it's questionable, if the client wants a no drug policy program, I say to them, let me send it off to a laboratory, and then I have it tested, for example, for marijuana. I've tested at 25 nanograms rather than 50. That way he knows and I know that was a true negative, at least down to that level of testing.

How to handle shy bladders. We can give them options of testing other specimens. They may want to be able to test hair. I have one client, not a client but a female, she couldn't pee for four hours. Let's face it, that's a real shy bladder. I said to the client, let's have her volunteer for a hair specimen. She agreed to it. We sent it off, positive for cocaine.

These are the types of things that the client who wants a drug free program relies on a collection site for. We are not just a deterrent. We are a detection, part of the process.

I also want to talk about the economics of drug testing, especially the on-site kits. I find that the prevalence factor is really the key. If the work site has a very high prevalence, I do not recommend to the client that they use on-site testing. It's not cost effective from my perspective.

That's my comments about drug testing and the sites.

MR. STEPHENSON: What is your prevalence threshold that you use on-site?

MR. FORTUNA: It sort of depends on my client because I charge different prices. If they only have like eight or ten employees, I charge one value. It is usually about 15 percent. If their prevalence rate is over 15 percent, it's not cost effective for me to suggest to my client that they use on-site testing. That's what I charge.

MR. STEPHENSON: Thank you.

MR. SHULTS: I think that's a real world view. I think that gets back to the philosophical issues. What's the thumb rule of marketing? Sell them what they want to buy. You don't want this, we will sell you this. We have this, we have a solution for you. If you don't want to use an MRO, we don't use an MRO. That basically I think crystallizes the problem. It crystallizes the problem, okay, you have a donor now. You have a donor who comes in and applies for a job. They are taking over the counter nose drops, the most common example. You get a screened positive. Now you say, you know, you got a positive here, you look like methamphetamine, but you may not be, but for $55, I can send this off to a certified laboratory, and if you are taking Vicks inhaler, maybe I will send you for another $20 to have an MRO verify that or we will do an D&L, that will cost another $250, what's wrong with that? The longer we do these programs, the longer we find that these issues, these isolated issues, show up. Keep in mind that not everybody is so enamored with drug testing, that the war on drugs will basically wash out these little safeguards, these little concerns, because it turns out that the safeguards that your employers don't want to buy saves their butt.

I think that is what it is all about. Your problem is not that you won't do well regardless if we have standards, what you will always be doing is competing against your competitor who will sell them what they want to buy. There are all kinds of new mechanisms, all kinds of things, things that I've only heard about just sort of conceptually, that will be new ways to evaluate substance abuse.

The question is are standards good and again, do they help or hurt. Of course, my argument historically has been we are talking about a very important issue here and that they do help. I don't like the idea that a parolee has more rights than a job applicant does.

MR. STEPHENSON: I just want to clarify one point for the group. We have asked Ted Shults to work with folks that would be interested in doing so on the issues of the roles of medical review officers on the alternative specimens and technologies. That doesn't mean it's going to be just his point of view, but it means he will chair a group of interested parties, professionals and practitioners with the experience in the real world, to sit together to help look at what we are going to need to do in terms of education materials, enhanced certification for medical review officers, the kinds of analysis of options that we need to look at in the Federal agency programs as we prepare our revised regulations.

There are certain things that we may not want to do that are perfectly acceptable to non-regulated private sector employers. I think it's important to realize that we are driving this program from a Federal agency and those that by extension use our rules and procedures. The rest of the world out there uses ours as kind of a bright line, kind of a guiding mark as to where they should try to position themselves. We are not always going to be convincing nor compelling, but we can still be present. If any of you have anything that you would like to contribute in terms of your own experience or to share, contact Ted. We want to work with large established certification programs that do training of medical review officers and again, get the experience from that real world point of view.

MR. EVANS: I would like to update you on what NOTA has been doing. NOTA has developed a set of draft standards for on-site drug testing and also on-site alcohol testing. The NOTA Board met last night. We met with Jeff Smith from DATIA. We have been talking with DATIA now for a couple of months about a joint training certification program of on-site test operators, and I think last night we came to an agreement that we are going to proceed on that joint venture between DATIA and NOTA, to develop those procedures and certification programs and standards. I will predict, if I can get it done, within six months, NOTA will come out with a set of standards for both drug and alcohol testing. We would like to work with you also on that. I see the MRO's being very critical in that process. I'd like to invite anybody that has any input, anything they would like to contribute to it, we would welcome your input in that regard.

MR. SHULTS: I recognized that would be a role that you would be doing, and there will probably be about 10 or 15 other groups following that as well, since there's no barrier to entry into that. I think that the issue there is what are the common elements, what are the things that everyone can buy into. Simply because you are going to do it doesn't mean that the marketplace -- you will either be large enough to do it or be able to do it. The question is how do you certify 20,000 or 30,000 sites and there will be a number of providers who will be doing that. What I'm looking at is what are those common elements that we need to do, not only for urine, but if we are talking about oral fluids as well. I appreciate your input and I know you have been very proactive at the state level. The other distinction is we are not really looking for legislative success as we are just marketplace success, you know, the idea of what's acceptable that will be a voluntary kind of approach to do this. I would encourage your organization also to re-visit its model policies to requiring certification of presumptive positives and medical review officer use.

MR. EVANS: We do require confirmation of positives. That's always been our position. I don't know if you had that misunderstanding. We have always recommended that.

MR. SHULTS: Some of your members have lobbied -- I apologize. I didn't want to attribute that to your organization, but some of your members to me are lobbying for we just need good on-site testing because drugs are so bad, that type of thing.

MR. EVANS: They may be selling on-site tests and they may or may not be our members. Let's not get those confused.

MR. SHULTS: Okay.

MR. EVANS: There are several manufacturers who are not our members. We have certain standards for our membership.

MR. SHULTS: Good.

MR. EVANS: That is contrary to our official position. We call for confirmation and MRO review and everything.

MR. SHULTS: MRO review is part of it?

MR. EVANS: Absolutely. Positive tests, we recommend it be confirmed and reviewed by an MRO. That has always been our position.

MR. SHULTS: Thanks. I look forward to working with you.

DR. WU (ACL): I would like to make a comment, both for Ted as well as for Dave. I don't know whether you are aware that the term "MRO" perhaps in this audience only meant the practitioner or physician whereas in the actual practice, many states, if you review the result, you are the MRO. This is especially true for the private industry testing. For example, the State of Oklahoma, you can be an MRO, even a chiropractor. Recently, the State of Iowa, my neighboring state, if you are a nurse, you can be an MRO. However, if you are a Ph.D. in toxicology, you cannot. I want the two of you to have some understanding, particularly when you talk about philosophical reasons, about economic reasons, please don't leave out the Ph.D. toxicologists, will you?

MR. SHULTS: Oklahoma passed a law when it required all results done in the state to be reviewed by a medical review officer. Interestingly enough, it raised an interesting issue for me because they authorized Ph.D. toxicologists to act as MRO's. We had to go back to our physicians and say, look, the state is authorizing this, and the state also said that you had to obtain training through AAMRO or MROCC. MROCC said we will only train doctors. We had to go to basically accommodate the Ph.D.'s, I'm actually not a physician, if anything, I'm more of a toxicologist, we said, okay, based upon that state law requirement, we would provide the training for them. Let me tell you something, nobody has been trained. There hasn't been a large interest.

It does raise the other interest of how drug testing can be shaped at state legislators more effectively because they are small and targeted. The laboratories are unduly influential at the state level. Some states, we see we are looking at alternative technologies. If you are based in that state and you have a lobbyist, you can basically say the only way you can do drug testing is with ear swabs, those types of things are going to happen out there. Again, I think there is a great deal of concern about expansion of this.

Again, from the perspective of is that appropriate, is it appropriate to have Ph.D.'s. I used to be the MRO at my laboratory before we even had MRO's. I was the attorney. I had to deal with these issues. Even though I have had medical training, I said to myself, I don't think this is appropriate. I don't think it is appropriate for me to be talking about private medical issues, not being at least in the health care profession. Again, those are issues. As a practical matter, those are peripheral issues, 99% of the MRO's out there are physicians. The bigger issue is some of them don't know what they are doing, which I think many in the lab business are well aware.

DR. SMITH: I have two questions and one comment. My question has to do, perhaps it's directed more at my former colleagues on the Federal side, as to whether or not it might be possible to at least look at designing a side by side look using Federal specimens, not at the point of collection, but at this point, since we have a repository at the laboratory, for doing some point of care, non- instrumented tests on specimens when they come into the laboratory and being able to then -- I know we have a lot of comparative data that's been done by the manufacturers, some that has been done by Bob and by Leo, but I am wondering if in terms of being able to look at Federal specimens, if that might give us some indications of a comparison from a science standpoint with regard to results for both presumptive negatives and positives.

That's my first question, whether that would be a possibility. I'm thinking back to the days that we have done that in other areas where we have been faced with taking a look at different policy and program directions, whether it be with regard to changes in cutoff levels, Donna, or diluted specimen definitions, et cetera. That is my first question, whether that would be possible.

The second comment is there have been throughout the past day and a half, discussions or comments made about comparison of looking at the policy for Federal testing for non-instrumented drug testing and comparing that or drawing parallels to the decision with the Department of Transportation for breath alcohol testing, which is essentially a point of collection, if you will, test. Two things. One is that Mr. Evans stated that the training requirements for that were by the manufacturer. I'd like to clarify that is not true, in fact, the DOT spent a lot of time to put together model curriculums. I would state that if in fact we are going forward with a type of delivery service, that we should again think of not putting the training or certification requirements in the hands of manufacturers, but making some attempt on the Federal level for Federal testing to identify model curriculum standards.

The second point that I would like to make relative to alcohol testing is when we get to the point of talking about non-instrumented drug testing and the policy and program decisions that will have to be made for federally regulated testing, there is a significant difference, and that is when you look at the testing of incumbent employees, whether that be random, reasonable suspicion or post-accident testing, with alcohol testing, you do have a confirmed result within 30 minutes. You do not have the issue of what do you do with an employee who has a result of "needs further testing" or presumptively positive that you cannot confirm for at least a day or two days, and the impact that has in terms of policy.

I think those are two significant differences from where we were four years ago with breath alcohol testing and some of the things that we need to consider from a federally mandated program perspective with non-instrumented drug tests.

My second question is to Ted in terms of his comments. Has it been your experience with medical review officers that in reality, their trust in the technology of non-instrumented drug tests is as good as the technology currently used for laboratory based screening analysis?

MR. SHULTS: I think physicians generally are skeptical about claims that are made. People come into their office all day claiming things. This will be better, that will be better, this drug works, that drug works. They are looking toward verification. The advice I give is trust but verify. To some degree, I think there may be a lot of uncertainty. Remember, these are a lot of physicians who -- the growth in the MRO practice has been from physicians who never really have been involved in these programs before. They have been trained. It is not something they learn in medical school. Their group now is in a managed care group that's offering occupational services and they get tapped on the shoulder, hey, we want you to be the MRO and we will go to this training program. They are coming into it relatively with a clean slate.

DR. SMITH: Do you suspect that the perception that the technology for non-instrumented drug testing devices for screening is less scientifically accurate and precise than for what is used in the laboratory based screening is because much of the printed material that you see compares the non-instrumented drug testing result to an ultimate GC/MS confirmation result, and therefore -- I'm not articulating that very well.

MR. SHULTS: I know what you are saying. I think to the lay person that looks at the type of data that we have been looking at for the last couple of days would say, these devices don't work. When I presented material, they say, order me false positives, false negatives. I have to step back and say look, we are talking about performance around the mean of atypical types of specimens in terms of what you get. I make it simple by saying let's just take 1,000 specimens, split them in half, send half to the laboratory and do half the other way and see what discrepancy you get.

DR. SMITH: I guess that comes back to my point about whether or not some kind of an absolute side by side parallel real kind of study using enough samples.

MR. SHULTS: Doctors are getting more and more used to point of care type tests, whether it's at the patient's bedside or with pregnancy tests, and they look at these things in sort of a beneficial light in a sense of I get my results very quickly, but there is a mental distinction. This is not a clinical diagnosis we are looking for. We are going to have some legal consequences from this result, so there is where there is more caution about the use of these devices. Despite the fact that 1,000 specimens may be reconciled, we do have the ongoing issue of who is watching the hen house, how do we maintain the quality control with these devices.

DR. LAPPE: One of the interesting things, Ted, I think, did a great job, the thing about standardization is essential, if you look back not only ten years to the DOT mandates, but go back to the time when there was no standard, if you remember, in the ten years preceding the DOT regulations, a radical departure from the way laboratory services were sold, when laboratories started selling their services not to physicians but directly to the workplace, and that was, I think, a revolutionary event, in carrying the message that drug testing is a decision that can be made by the employer as to how the process is done. When you de-centralize and unbundle the process, it goes all the way back to the employer, by centralizing and bundling the process, as the DOT did in 1989, we moved it forward from probably less than 10 million tests a year to close to 50 million drug tests a year today, and I think we are going to approach 100 million drug tests in the next ten years, and by unbundling and de-centralizing it, the employer is going to pick and choose what they want. It's not even going to get to the level of the collector, like Mr. Fortuna said. The decision is going to be made at the workplace, that for some reason, they know more than the collectors and the laboratories and the MRO's about what they want.

DR. WILLETTE: This afternoon, we are going to have three presentations to talk about oral fluid testing. Oral fluid testing is nothing new, certainly when it comes to alcohol. I was reminiscing about in the 1970's when I was at NIDA, we had a large program with the Department of Transportation, with NTSA, looking at ways of detecting drunk drivers. There was a lot of work done at the time on saliva. Unfortunately, the technology wasn't available to make that readily available road side or even at the station house. I think it's kind of exciting that we have people here to speak about the realistic prospects of doing that.

Our first presentation is Sam Niedbala from STC, who is going to talk about what they are doing or where they are at with oral fluid testing.

MANUFACTURERS AND DISTRIBUTORS PRESENTATIONS ON ORAL FLUID TESTING

PRESENTATION ON BEHALF OF STC TECHNOLOGIES

DR. NIEDBALA: It has been interesting listening to all of the discussions yesterday afternoon on urine testing. Let me preface what I'm about to present. For those of you who attend these meetings regularly, you have seen me rush in lately to present the updates on oral fluid, and the term "oral fluid," actually the other working group has agreed to use rather than "saliva" for clinical reasons, because saliva and oral fluid are really two different things. It's kind of a joke going around, whether or not you say saliva or oral fluid, but there is a rationale for that.

Today, for the first time I'm representing a new technology that we have been working on, which is different than our lab based approach using an oral fluid specimen which is shipped back to a laboratory for analysis. This truly is an on-site test, driven by interest in both avenues of testing.

I will try and do my presentation based on the questions that have been asked, but I will tell you right up front, this is not a technology where at the end of the session today, you can walk up, give me your cards and I'm going to ship you devices tomorrow. This is something that will be introduced to market next year, but we feel it is important to come out of the closet on this because as the regulations are being developed, you really need to know what's on the near term horizon technologically.

The testing principle that we use is based on antibody. It is immunoassay. However, it utilizes UPT or up converting phosphor technology. A little bit of background on this. This has been also a topic of discussion in funding here in Washington because for the last several years, DARPA, the Defense Advanced Research Projects Agency, has funded over $10 million worth of research to develop this for a chemical and biological warfare defense. It's currently heading out of the laboratory and into field tests next year for things like Anthrax, Ricin, Plague, all of the situations which we may not talk about in this group, but it is a real potential problem for society at large. What STC had done is several years ago licensed this technology from several organizations and has been working to develop it into a base platform, drug testing being one of the first applications. Up converting phosphor technology. What it is is a patented technology that is only one part of any particular testing system, whether you use antibody, DNA probes, and of course, this does require an inexpensive instrument to detect, and I'm going to show you that.

You are familiar with lateral flow technology and what we have utilized here is the UPT technology inside this platform. This is not the final drawing of what we are doing here because we will be able to screen for multiple tests simultaneously, but what we have is the same constraints. In other words, there's practices and habits that people have already developed for use in these lateral flow techniques, and we are not changing that, but rather we are adding the features and benefits that come as a basis of the technology. Let me explain what up converting phosphor technology is. In another form, it has been around for actually several decades, inside fluorescent light bulbs, televisions, high density displays, there's a coating of particles that exist on those screens. When light hits that, usually ultraviolet light, those particles glow, so we see white light in this particular case, visible light. Also on our television, we see multiple colors. Those particles, the same basic shell of a particle, are what's now incorporated in this particular technology. Close your eyes for a moment and think through history. You have enzyme immunoassays, radioimmunoassays, fluorescent immunoassay. Now what you think about using as the amplifier for the immunoassay are these up converting phosphor particles. This is not an in-depth scientific presentation of it, but because of these particles, you get several key benefits that are really not available with other materials that you may use as reporters.

Let me try to put this into an analogy. Multiple colors are possible with the technology in a single detection of a strip. The analogy to this is to think of your television. You can turn it on and simultaneously see multiple colors with a single light source that is exciting all of those different materials. The difference in what we do here versus your television or fluorescent light is we use infrared light, which is the source to excite the particles, and they emit light into visible wavelengths. If you go back to your old chemistry that's called an anti-stoke shift. How many of you remember that term? Anti-stoke shifts don't exist in nature, so no matter what matrix you put in there, whole blood, whatever it is you grind up, environmental samples, there is no background, and the number one killer of immunoassays is the background or the non-specific binding of materials. Consequently, we get the third result, which is extremely high sensitivity with these materials. In addition to that, they become a permanent record, just like your television which can be turned on or off or the light on or off as many times as you want, these materials have the same characteristics.

Supports miniaturization. These strips, and close your mind and think about how people set up the architecture of a lateral flow strip, it's generally about the size of your pinkie, which is somewhere between five and seven millimeters, but we literally shrink that down to one millimeter so we can miniaturize the strip size, meaning we can also drive some of the sensitivities on these materials because we can put more fluid, more sample through the strip.

Interference free. The fact that it up converts. I'll show you a whole list of materials that have been tested without any interference. That doesn't mean that if I put something caustic which would fry the antibodies, as an example, it wouldn't affect it, but the label itself is not interfered with as you can have with say gold or other materials. This is important. Even if you think about oral fluids, one of the questions that Barry should be asking me is what happens if somebody has bad oral hygiene and their gums are bleeding and I put that specimen into the device. What happens? With this technology, absolutely nothing. That's where you get to the last part of this, which is suitable for any testing matrix.

The overview of the way the system would work is really not that much different from an urine test, but you do have one other component here in the up front. You have to collect the sample somehow. I'm going to go through the pieces, and then on the end, you have a small little inexpensive instrument that is used to read the end result. That instrument, as far as our philosophical approach to this, and I've heard this over and over in the last day and a half, is the information generated is the key point and what you do with that information, who looks at the information, what the interpretation of that information becomes is really at the end of the day what we are talking about. The immunoassay is just one small step that has to occur to get to that information. We have built in an instrument with the software capabilities to either share that, if that's an MRO looking at it, or as the end user finishes there and simply needs to be stored in the future.

Let me talk through a couple of these pieces. The UPT reader. This is about the size of a palm pilot, so it fits in your hand. This is actually pretty close to what the version is that the military is evaluating as well. The feature I want to point out is we have made this fully compatible to communicate the information, whether that is through an IR port, an IR 232 or direct modem connection. Any of those things are possible. We are trying to take what we have learned in the past, being the company that had the first saliva alcohol test, and having gone through all of the development of the training programs for that in addition to the technology, and build in what we would like to have improved way back when into this particular device.

The collector is actually -- in essence, in the room, and I'll disclose more of this to the Board, some of my competitors. There are a few things in here I'm not going to show you today.

What we have designed in this is actually that the collector is easy to use, you put it in your mouth. There is a sample adequacy built into this. Again, one of the things we would like to improve from the past, so that the collection volume is fixed and you know when you have collected the right amount of specimen.

The test device itself, easy features or features are easy to hold. There is a little instrument interlock that's in the design of the cassette that's on the device so that when it's put into the reader, that's the only time the reader will turn on and recognize it. We tried to build in three to four levels of quality control and user interface to make sure that everything is functioning correctly with this device. There is also a portion of this where when you collect that stick and you put it into the device, it cannot be pulled back out, for those of you who have used our alcohol tests, it's one of the things I hate about that, you put it in and then the stick falls out, and you haven't properly filled it. We redesigned this so that when it goes in, it doesn't come back out. Picture it. The stick is collected. It has a sample adequacy. It goes into the test cassette, locks in. The sample runs up. You can actually run the sample one of two ways.

You could do it in a batch mode where you would simply wind them up and put the sticks into the device and let it run up the strips. In that case what happens is there is also a visual QC window, and that's to let the operator know that it's safe to put it into the instrument, that the test has been performed correctly or conversely, you can put it in the instrument and let it count it for you, take the end result, and you simply look at the end product.

This is some of the assay sensitivity data that we have generated using oral fluid specimens. The only one that's different, and I just like to put it up here, is LSD. We have actually done sensitivities in that range in urine. For the first time, what we are seeing is a technology that in oral fluids is demonstrating the required sensitivities to have the greatest window of detection.

Let me show you some of the data. These data are actual scans of some of the strips that have been used to run the immunoassays. The user wouldn't see this result. I'm putting it up here so you guys can get an idea of how the test works. What happens in this test -- you brought up the question yesterday, is there an immunological control in the device, and the answer in this one is yes. In this particular case, the first peak that you will see is a control, which is immunological, looking for an anti-species antibody, which makes sure, one, that the pH was correct, two, that it filled correctly, and three, that the biological's are functioning, and the second peak that you see are the data from the actual tests.

This is BE, this is .1 nanograms per mL, 1 nanogram per mL, 10 nanograms per mL and 100 nanograms per mL. What you notice just very simply is that the peak is decreasing. What this is is a laser scan across the wick so that you can see the increase in absorption that results from these up converting phosphor particles. Notice that the background here is just about nothing, and that is exactly what it is. If I represented this numerically, these are well above 1,000 counts as an example, and that background that you would see would be about five to ten of those units. If you plot that, you can see the displacement curve, classic immunoassay, and you can get an idea of the range we are working in. The instrument itself is capable of seeing about five log orders with these phosphor particles. However, any immunoassay is always limited by the antibodies in terms of what it actually would produce.

Here's one for THC, a little bit different. This is 1 nanogram, 10 nanograms and 100 nanograms, but the same basic idea where the immunoassay displaces as you are used to now, except with the up converting phosphor particles. I'm missing part of my curve over here or plotted here. You can see the representation of the assay.

Here's one for LSD. This is just the LSD test spot itself. You can see with no drug, .25 nanograms per mL or 250 picograms per mL, .5 and then 5 nanograms per mL, to give you an idea of how the technology works.

The next one is an example of multiplexing. Just for fun, we took benzoylecgonine and LSD and tried to take a look at how this would perform in a multiplexed assay. You may have some difficulty seeing this from the back of the room, so let me just explain what's going on here.

This particle is a blue particle. This particle is a green particle. This one with a blue color is for cocaine. This one is for LSD. In the same strip with a test run with no drug present, these are the zero values that you get for each one of the drugs. If I add in 500 picograms per mL of LSD, you can see this curve decreases down. This curve stays basically the same. If I now run another strip and I add in 10 nanograms per mL of BE, you can see this curve go down from here from the first two, where the LSD comes back up. Part of this is to show the example of how different colors are spectrally unique with this. Again, in 15 minutes, I can't go into all of the technical details, but any one of these colors can be easily seen and distinguished from one another, so you don't have the potential, unlike colloidal gold, where if you get non- specific binding of the colloidal gold throughout, this one distinguishes based on the particular color particle that has been tagged, so you get the specificity even in multiplexing.

What we believe is going to be the stage for rapid saliva based tests is really the same as the laboratory based tests. These are the cutoffs we are working against right now in the lab based system. As you can see, this is going to match easily what we do between the two usage situations. To answer a few more of the questions that were given to us to answer. What specimen treatment or handling is required? These specimens are untreated. You collect the saliva directly and deliver it into the device directly without any sort of pre-treatment.

Is the product FDA cleared? No. They are not FDA cleared yet but they will be. That's just part of the development cycle for this. We also have the predicate devices which we have done for lab based testing, so we really see that as just a very natural extension.

How do you recommend dealing with adulterants? As I said before, the UPT technology is really not affected by materials in the solutions. Again, we have tested all of these materials, which include cells, buffers, chelates, just about everything we can think of to throw in there. That doesn't affect the label but I have to also caveat that if we throw in caustic whatever, we will fry the antibodies. There are those situations. Of course, in our other group, we are having discussions about what is a contaminant in oral fluids, and the way we are handling that in the other working group is to say, look, there's going to be a waiting period. There is a clearance that is then allowed to occur in the oral cavity. I think this is where the synergies will occur between the two groups, at least we hope, and the way we will treat the interference testing.

How should the device be evaluated prior to use? Each strip is self contained, much like we do now with our alcohol tests. Within the strip, and again, I said I didn't put up all the details of the tricks we have, but each strip will self calibrate. This is a divergence from what other people have done in the past, and I think to answer some of the questions and concerns I've heard in the last day and a half, there is a composition of these phosphor particles different from all others that is on each and every test strip. The instrument indexes on that and uses that as a point not only to do QC, but also can quantitate because of that particular attribute in the system as it is being designed. We really see several levels of quality control. I am going to talk through that here in just a moment.

What are the unique features of the product? I'm trying to summarize these as best I could. Ultra sensitive. Rapid. All of the tests that I've shown and all of the data we have done are five to ten minute tests. There is nothing outside what you are used to now, with the potential to speed that up even more. They are quantitative, although for this application, we expect them to be qualitative. We don't see them affected by interference that affects the colloidal gold assays, and it is a direct oral fluid sampling and use.

In our other group, we are talking about taking oral fluid specimens and having to process them through another collector into a laboratory. This is a direct sampling method.

What are the quality control recommendations for the device? The instrument, as I said, has several different levels of control. It looks for the immunological control to make sure the test worked properly. It has a reader check for each test. That was the index against a phosphor composition on each strip, and then also you saw that we built in a visual QC so that you also can ask the operator what they have seen and make sure that before it went into the instrument, if it was done in say a batch mode, you knew it filled properly. The instrument will also record that once it goes inside.

What are the recommendations for training on the use of the device? We did, Donna, this is where I disagree with your statement earlier this morning about DOT developed the training guidelines, we did work with DOT very closely to develop the training guidelines for saliva alcohol testing several years ago, and we envision the same sort of thing here. Nowadays we have not only video and print but we have Internet, and you have heard some of the other innovative ways that people are doing training. I think all things will be possible. I also agree there has to be control and standardization. When we did the first STC training and then trained the trainers and worked very diligently to try to make that a quality program, it wasn't because we wanted to slant it towards the manufacturer, and I've heard a little bit of slamming going on here about those kinds of things, it was because we wanted to make sure that one, the science of the screening matched up with what the training needed to achieve. That, we expect to do with this technology as well.

How do you recommend using the device in workplace drug testing settings? This is a rather open question but without putting up another algorithm which everybody has seen over and over, I just rather point out a few things. One, that we see that the specimen adequacy now is built into this device. We have blocked out as many errors as we could see between the interface of the person operating or doing the test and when that goes into the instrument. The second thing we did is that the read out or the information generated has to come from that instrument itself. It doesn't come from someone visually looking at this. You can't see these particles with your eye. The third thing is that the information that's generated is stored. This particular device I showed you will be capable of storing up to 5,000 results on board. In addition to that, that information could be transferred, monitored. Again, philosophically, we believe control and oversight is a part of this, because there will be mistakes and we design all our government so we have checks and balances, why shouldn't we do it here with drug testing as well.

Finally, do you manufacture the products. Yes. All tests are developed and manufactured by STC. The instrument is done for us by a corporate partner. There are other uses for the technology which these folks are doing, but in exchange, they also will be supplying these instruments to us.

As far as getting it out to the marketplace, we will expect that it's not only us, but we will have distribution partners to really make this, we hope, a success in the marketplace.

DR. WILLETTE: Our next speaker is Vina Spiehler. She's speaking to us on behalf of Cozart Bioscience of Abingdon, Great Britain.

PRESENTATION ON BEHALF OF COZART BIOSCIENCE

DR. SPIEHLER: I have slides to show. There is the first one. I have been asked to present this on behalf of Cozart Bioscience Limited in Abingdon, U.K. They produce a saliva test system which consists of a collection device, a lateral flow immunoassay with gold labeled antibodies, similar to all the ones which we were shown yesterday, and a reader. This is the picture of the reader. The reader is about the size and configuration of a mobile phone. That is in part because the test was developed with U.K. Government money, in conjunction with three regional police departments, and that happened to be an ergonomic forum that the police knew how to handle and could handle. It contains a screen so that the results are read off of the screen. These are drug abbreviations. OPI, for instance, is opiates. There is the word "negative" in this case. If all results are negative, this little light here will shine green. If any of these results are positive, this light will be red. This is just a start switch here.

You can see that the cassette fits into the instrument. It also has an interlock so the instrument can tell when the cassette is in place and in the right orientation.

In addition, the reader can download. It's a little computer chip inside. It can store and it can download results, so it could put them through a modem or an infrared port or a cable into a computer system, and I suppose there is nothing that would prevent those results from being downloaded on line into an MRO system, like we were shown yesterday. I don't know of anyone doing that.

RapiScan right now is being sold in five, four and three panel configurations. For instance, the five panel would be the amphetamines, benzodiazepines, cocaine, cannabinoids, opiates. There is an immunoassay positive control spot and a negative control, the reader checks that the background has cleared. The sample is collected and the collection is part of the total test system. It is collected on a pad. I have an example here. It is placed in the mouth, absorbs 1 mL. We know it absorbs one mL of saliva because there is a little indicator in the handle that turns blue when one mL is collected. That one mL of saliva is added to two mLs of buffer in a test tube. The pad is removed from the handle and the buffer and saliva are mixed by flicking the tube, and then the pad is compressed by putting in this plunger. It looks to me like a serum separator plunger. Finally, the mixture of fluid is pipetted using a disposable pipetter, which is also sealed in the package, into the cassette. It doesn't take all three mLs in the cassette. It in fact takes about 250 microliters to run the cassette, so there is some fluid left after the test.

It takes the same amount to run all five tests or three tests. People using the three test cassette have been people in drug treatment programs where they want to look for opiates, cocaine and methadone, for instance. Most of the people are looking at the five test and the five tests that are selected are those that are involved in driving under the influence.

The cutoffs were set as cutoffs in the saliva. Because we have a threefold solution here, those are going to be three times the cassette cutoffs, which I'll be showing you in a minute. For instance, a 10 nanogram per mL 6-MAM cutoff on the cassette and in the reader would be equivalent to a 30 nanogram per mL 6-MAM in saliva. Since there is reported to be a sixfold amplification of 6-MAM from saliva to plasma, that is the saliva to plasma ratio for 6-MAM, that would correspond to 5 nanograms per mL in plasma.

To show you the performance characteristics for the device, the cutoff is morphine. Morphine is stable and is used. The cutoff on the cassette is 10 nanograms per milliliter for the opiates, but it has equal cross reactivity with many other opiates. For cocaine, the calibrator is benzoylecgonine. It is set to 10 nanograms per mL in the cassette and it has about 20 percent cross reactivity with the cocaine. Amphetamines, it is a d-amphetamine specific test. It will pick up MDA, but is not going to be picking up what I consider realistic concentrations of methamphetamine or Ecstasy in the saliva test using a 10 nanogram per mL cutoff on the cassette.

Temazepam is the calibrator at a 100 nanograms per mL cutoff for the benzodiazepines with lesser cross activity to the other benzodiazepines. For THC, THC acid is the calibrator at 20 nanograms per mL, with about a 10 percent cross reactivity with THC, which is what is actually found in saliva. It's not quite as good as I would have wanted to see for the cannabinoids.

This I'm going to put in. This is my own recommendations of cutoffs that I suggest for screening saliva for drugs of abuse when the objective is to determine if someone is under the influence of a drug for purposes of driving. This test was developed with the police. It's being sold and the people I know who are using it are police forces and specifically as a roadside test to determine if someone is under the influence of a drug. For these different drug classes using the plasma ranges from users for a physiological effect, an effect on performance of these drugs, and a saliva/plasma ratio from the literature, I said these as suggested targets. I think the THC should be 100 and I'm assured by the end of the year that the Cozart device will be meeting that requirement for driving under the influence.

We have a very bad saliva/plasma ratio for Diazepam and the other benzodiazepines, so you would really need to get a saliva cutoff down to three or four nanograms per mL. If the device is positive for benzodiazepines in these drivers, it probably means they have used it recently and you are looking at residual drug in the mouth.

I think the device does have cross reactivity with analytes found in saliva at cutoffs that we have set initially to correspond to blood levels for DUI or toxic effects. The assay is an immunoassay. Their policy is that positive specimens must be confirmed and then finally to estimate blood concentrations of drugs, the saliva/plasma ratio must be corrected for the pH of saliva, which is not currently done with the RapiScan instrument.

Is the product FDA cleared? No, it is not. It's not cleared by FDA and it's not marketed in the United States. In England, the policy is post-market surveillance rather than pre-market approval, and it is apparently on their radar because the product has just been awarded a special millennial certificate for innovative cleverness. In fact, it is in the running to be one of the 50 millennial products that will be displayed at the millennial dome in Greenwich. The British are quite proud of their achievements with this product and the government of course helped fund it.

There are clinical studies that demonstrate product efficacy. The clinical studies have been carried out at U.K. universities and field evaluations by the regional police forces in the U.K. Some of those were reported at the SOFT TIAFT meeting in Albuquerque last year, more will be reported at the American Academy of Forensic Scientists' meeting in Reno. They missed the SOFT abstract deadline. I know there are field evaluations currently underway by police forces in Finland, in Australia, and in California by law enforcement organizations. There are also clinical studies underway with this product in the United States.

How do you recommend dealing with adulterants. The manufacturer recommends a ten minute observation period in which the donor does not smoke, consume food or drink before a collection of saliva sample. This is similar to that which would be carried out before doing a breath alcohol. In fact, the target user, the police forces, the roadside officer that is stopping a potential drunk driver, is used to doing all these things for breath alcohol, so the RapiScan follows that same model. My own experience is a person cannot hold their saliva, particularly if they have a solid or liquid adulterant in it in their mouth for more than a couple of minutes because that stimulates production of saliva and you either swallow or spit or dribble it down the front of you. I think a ten minute observation period is probably adequate.

At the start of each day, a battery check is done on the reader and the RapiScan instrument calibration is checked using a positive/negative system test cassette. These are printed cassettes. They recommend that with each new lot, you run a genuine positive and negative mixture. Most users just use their own saliva for the negative and the manufacturer has a mixture that they use themselves for quality control checking for positives, which is set above the cutoff. I don't know how far above. I don't know what the threshold profile is of the instrument, so I can't tell whether it would perform at plus or minus 25% or 50%. It might perform at plus or minus 25%. I do know what the optic characteristics are and it's possible that the reader could do that. I don't know if the colloidal gold and antibodies will perform at that level.

Unique features of the product are the on-site instrumentation which was developed to be used by a non-scientific person, a police officer. Often this police officer is somebody who is specially trained. Many of these instruments are being placed with the same officers who are being trained as drug recognition experts. They can do the DRE. They not only see driving which catches their attention, they also can administer field sobriety tests and do the different signs and symptoms observations of the DRE. This would be one more part of a DRE exam.

The quality control recommendations for the device, I have already mentioned, that the device contains internal quality control checks. It checks the operational control for each cassette for blank reading, for the immunoassay reaction occurring demonstrating the antibody at least has gotten to the proper end of the cassette. At the start of each day, a battery check is done and the RapiScan instrument is calibrated or the calibration is actually checked using the positive and negative system test cassettes. The manufacturer recommends when a new lot of cassettes is open for use, a representative cassette should be tested using a negative saliva sample and another representative cassette should be tested with a known drug mixture with concentrations that exceed the cutoffs.

Collectors and testers should participate in any available saliva PT survey, and the training that has been provided so far has been from the manufacturer and with the distributors. There are distributors for different countries. Usually that training is similar to whatever that police agency is currently giving their DRE and their alcohol roadside testing officers.

One of the distributors, the Forensic Science Service of the home office in the U.K. is contemplating using the device in the U.K. in workplace testing settings. They have told me that they believe if there is a positive result, they will then have the collector take a second sample, a second saliva sample. Currently when they test for breath alcohol, the second sample, which is sent into the laboratory in Wales, is urine. They take the standard 30 minute time urine collection. At least in that application, the second sample will include the remainder because you only used a small portion of the total fluid, saliva and buffer, that is in the tube. The tube can be capped. A evidence seal can be put over it and sent into the lab. In that particular case, the people administering the program worked out a protocol where they can take a second sample, not necessarily a second saliva sample, and if it's a police agency, they may take a blood sample. For instance, the Australian distributor is running urgent care clinics and has a central pathology lab, so they have phlebotomists present. A positive saliva might result in a blood sample in that case.

The manufacturer states on their package insert that it is an initial immunoassay test and that any positive must be confirmed. They say GC/MS test is the preferred method. It also says clinical consideration and professional judgment should be applied to any drug of abuse test result.

I think I answered all the questions. The cassettes are manufactured at Cozart. I have been there and I have watched them make them. The reader is manufactured in the U.K. at a separate site.

The training is being carried out currently by the manufacturer and the distributor and the agency. In most cases, they are police agencies. Many times they are national police agencies. It seems to be similar to that that the agency is using for their roadside alcohol training and policies for certification for those officers. It seems to be often in conjunction with the DRE, the drug recognition exam training for those officers.

MR. FORTUNA: Is there any data that they have from using it in Great Britain?

DR. SPIEHLER: Yes. The initial studies which were done in the U.K. were reported at the SOFT TIAFT meeting by Claire Malcolm from the University of Glascow, and the results from the University at Oxford will be presented at the American Academy of Forensic Scientists' meeting in Reno.

DR. SALAMONE: Once the saliva sample is taken, how long does it have to elute into the elutting buffer? Is there some waiting period?

DR. SPIEHLER: No. There is no waiting period stated. You have to flick it until that pad leaves the plastic collector. It's perforated and it comes off pretty easy. That has to be done. It's easier when wet than when dry. It probably takes a minute or two. Once you have pipetted it into the cassette, you are supposed to watch and see that the fluid is beginning to flow in the cassette. I forgot to mention that the reader has an incubation timer which is two minutes if it's a single test, but in the five test cassette, it takes ten minutes of incubation of flow before the reading is made.

DR. SALAMONE: What wavelength does the reader take its measurements at?

DR. SPIEHLER: I really don't know.

DR. GOOD: Are there any recommendations for external PT on this system?

DR. SPIEHLER: The manufacturer encourages people to do external PT, whatever is available to them, but I don't know that they have set any standards for that.

DR. WILLETTE: Dr. Carl Good with Avitar will bring us up to date on their product development.

PRESENTATION ON BEHALF OF AVITAR

DR. GOOD: I'd like to say that we appreciate this opportunity very much and also to have been able to participate in the Subcommittee on Oral Fluid Testing that Sam Niedbala has chaired.

I think this is a very interesting time with regard to drug testing and many analytical procedures in general in the sense that there is a lot of new technology coming to the forefront. I think much of the technology can be very beneficial to the end user, but it's a question of simulations, setting the proper standards and having people familiar with the different types of systems that are becoming available.

I'd like to tell you a little bit about Avitar Corporation itself. We are not exactly a household name but we do have a lot of experience in the medical device area. Our chairman is Pete Phildius who ran most of Baxter Travanal, also National Medical Care and several other major manufacturers in the medical device area and medical procedures. Doug Scott, our president, did the international manufacturing for Baxter, actually did a stint for a while as head of quality control at Frito-Lay and has worked in other medical device companies. We have a group in middle management that also are very experienced in the medical device area. Avitar is ISO 9000 FDA licensed GMP facility in Canton, Massachusetts. CE Mark. I have EN46001. I'm not exactly sure what it is but it is some type of European control. We recently have merged with U.S. Drug Testing Laboratories in Chicago, specialized in hair and other fluid drug testing.

Avitar's original purpose was to exploit some technology in the area of wound care. This was a foam that was very soft but also had high absorptivity of liquid, and in essence, would take in about 17 times its weight in liquid, but it turned out that this same foam, because of its chemical backbone, is very low in the binding of many, many different analytes. In fact, in some cases, it can concentrate them a little bit, especially with proteins, because it binds water more tightly than the protein. It came to our attention from some people who were looking in the area of rapid testing that this would make an excellent oral fluid collector and is currently used for DNA sampling of the buccal cavity but also has been used for HIV testing in some products in that area. A patent was submitted and Avitar is the exclusive licensee of a patent using this type and other foams for oral fluid collectors.

In looking at the different opportunities, Avitar decided that the drug testing area and oral fluid testing was something that provided a lot of promise and might fit well with its capabilities in medical device manufacturing. For a good oral fluid test that is a rapid test, it is much better if you add low absorptivity to the collector because then you can use the sample itself without the user manipulation of having to extract the material you are trying to analyze from the collector itself. In a rapid test device, you also need good flow of the sample because it's difficult in some of the devices because of the viscoelastic nature, you might say, of saliva or oral fluid, to get it to flow way through a system. There are things that we have done that have been submitted for patent approval that have to do with having oral fluid flow effectively in the device in a wide range of sample types, and then appropriate system sensitivity. This is a challenge to people coming into this field because there really are no guidelines yet as to what the sensitivity of your system should be. That's why I think participating in this group has been a big advantage and in some international groups, too, because there is somewhat of a consensus developing I think about the cutoff levels for the more important drugs.

For our collector which is shown up here, again, this is a patented device, so the foam and its use as a collector has a patent on it, but also the configuration of this type of collector is also patented. We have low binding of analyte, it's easy to remove the analyte or the sample from the collector. It doesn't require an extraction process. The material is non- toxic with proven oral safety because it is used in medical devices. There is a lot of animal data. It is also used in dental artifices as a foam mouthpiece for tooth whitening and fluorite treatment. You are dealing with something that is really medical grade to be putting in your mouth. Also, it's taste free. Some of the other collectors have salts and things in them that don't taste good.

We think that the benefits of oral fluid drug testing are you have a definite sample source because you can watch sample acquisition. One of the things I like to tell is that one time coming back from the West Coast at 3,000 feet over North Dakota, I gave the guy in the seat next to me on the plane a drug test, and I don't think you would do that with your typical urine test system. You can really watch the sample acquisition and determine where the sample came from.

I think at one of the other DTAB meetings, someone was saying if you asked people if they observe an urine collection, they will say yes. If you ask them do you actually watch the urine come out of the urethra, they will say no. This is a situation where you can really watch the collector come out of someone's mouth. You can get the results in minutes. There is no chance of sample cross contamination because you are applying the sample to the test device at that point in time. Confirmatory testing is a little bit of a different issue, certainly on the initial screening situation, that isn't an issue. It can be used in public situations. You don't require a special laboratory. It's easy to use. There is a reduced chance of sample adulteration, no special facilities and the results are in minutes for a rapid re-test if you want it.

This hasn't been talked about a lot, but if I were being tested, I would like to have the opportunity to challenge it at that point in time, rather than having to wait and come back at some later time, even with a split sample, to determine that I would want to contest the results. I just threw this in. I was going through the Web one day and picked up adulterants that are available. Several times at this meeting, the urinator has been described. It's a little pump with a bag of urine. I've heard that some collection sites actually have a metal detector trying to detect this. We have things like Fast Flush. I think there is a special shampoo for hair testing. This is estimated to be about a $40 million market today, in just providing adulterants to people in drug test avoidance or positive avoidance procedures.

In the operation of our device, one pulls the hood back on the collector. We usually tell people to develop a little oral fluid in their mouth beforehand, then put the collector in your mouth. Again, it's a very absorbent material but very soft, move it around for a minute or two, and then push the hood forward a little bit to start the liquid flow out of the collector, and then in a sequential fashion, to add four drops to the test system. The way I am showing this, the test reads the same way as a regular urine test. You would have a line that's missing on a positive and if the lines are there, it's negative. As again was described earlier for some of the urine tests, it appears that the device is effective if you have a short read time when you are dealing with a negative sample because all the bands come up. We have never seen a band disappear once it appears, but we recommend waiting 15 minutes for positives to in essence give every chance for a band to appear.

The interpretation of results, again, this isn't an MRO interpretation. Just reading the test is similar to the urine test.

There were a number of questions that were posed and I am going to try to answer them as best I can.

The immunoassay test principle is lateral flow immunoassay. The cutoffs on our initial tests are 10 for cocaine and opiates and 30 for THC. Specimen treatment. We don't require any specimen treatment. We don't currently have FDA approval but we are certainly in that process, yet this is a modification of devices using certain antibodies that have been in FDA approved materials which really relates to cross reactivity and inhibition of the test by different drugs or materials that might be in the sample.

Adulterants. We believe that they are reduced compared to urine, and because a lot of the adulterants in urine you wouldn't want to place in the oral cavity. In addition, as Vina described, if you watch the individual for ten minutes and give them a chance to clear things out of their mouth, it's never perfect, but you have some assurance that you have a reduced chance of adulterants.

For the evaluation of the device, I think there has been a lot of discussion today here about what to do for on-site urine testing, and we would assume the same thing would apply for oral fluid. We are going to be providing procedural controls, positive and negative, in a stabilized matrix simulating oral fluid so the customers can use them for their daily or weekly or batch-wise performance testing.

Unique features of the device. I think I have gone over that earlier. It's oral fluid, rapid test. There is no extraction. It's really a simpler procedure. It's multi-test. We recommend a procedural control use and the external PT, again, I think it's the industry and regulatory organizations that define what is required, and we certainly would go along with that.

Training, again, we are using video, on-site training, different procedures. We haven't implemented all the material on our Web page but will in the future, and we will have a certification program.

For workplace use, I think we will have the controls that we described before, and we are working on a reader automated system, not unlike the things that have been described here that would allow what can be a qualitative visual test to be read by an instrument so that you take the human variable in that aspect out of the system.

We are called a combined manufacturer. Some of the materials we manufacture ourselves, others we buy in OEM, but our goal in the long run is to be a completely fully integrated manufacturer of all the critical components in the device.

In some of our trials, we have tried the device system in several different areas. I want to talk a little bit about use in a methadone clinic. One of the reasons for this is if you look at people entering the methadone clinic, you have admitted drug users who we have found are fairly open in describing the types of drugs they use, when they took them, the amount of drug, because they are in there to be helped. We have done some studies at methadone clinics especially in the Boston area, where the folks have reported what drugs they have used and the time they have used. There have been studies in the literature that are really dosage studies, where individuals are dosed with the drug and then you look at how long that drug will appear in a certain body fluid. I think these studies are very good for determining the distribution of drug in different body fluids, but probably aren't quite as useful for determining the availability of the drug in time, because the dosages used are fairly small and atypical.

We have accumulated with this group information on what other non-abused drugs they have used, what abused drugs they have used, and the food they used. We have gotten matched samples in urine. We have gone through the whole process with GC/MS. We have oral fluid and both performed in the rapid test and with GC/MS. Here is just a list of some of the different things people have had in their mouth. Watermelon was a big one this summer and just what you would expect. I think the important thing is that you really don't see any type of effect on the test due to these materials. Some of this information is anecdotal. People tell us what they have had, but we are also doing studies where we are having people ingest certain materials and then look at its effect on the test run spike samples. As far as the Beluga Caviar trials and the hot fudge sundae trials, we are having no problem. The brussels sprouts trials were a little bit of a difficulty. If anyone would like to volunteer for that, we are looking for volunteers. This is just a list of some of the other drugs that have been taken by people who tested positive on the test. Again, this is from the clinical trial data. It is somewhat anecdotal, but we found in general, people are pretty honest with us. We have done work on the interference of drugs in the assays as one would normally do for a 510(k) submission.

Since we do urine GC/MS and saliva GC/MS, we can check just how truthful these folks are. In general, in this situation where they are folks going into a clinic and that are seeking treatment, we have found them to be relatively truthful. In general what we found is that for opiates in the sensitivity we have described, if we look back the day they have taken the drug and a day later, you get a sensitivity and specificity in the 94 to 95% range. Cocaine was a little bit less for the day on which they took cocaine or within the first 24 hours, it was pretty much 100%. It dropped off some in the second day. This is a comparison to the urine test run by the EMIT system, and if you look at it further back in time, some people tell you I took the drug on Thursday and you test them on Saturday, then you get more of a disparity between the two, but in this process, we have detected people in oral fluid as being positive let's say up to four days after their last admitted use. THC, with the sensitivity that we have, it is just back a few hours. I think in the future, we are going to want to push the THC down to a higher level of sensitivity. That is really a current use situation.

We think the benefits of an oral screen is it is a simple procedure, no extractions. We believe we have appropriate sensitivity. The THC, we will probably bring down in the future. It's certainly more of an under the influence indicator. This is an issue. You might have these tests used different ways for different situations because you may want more of an under the influence determination so you would have a higher cutoff and you may want to look back further and you would have a lower cutoff. The results are visible, so you don't have any calculations or numbers, but it can be instrumented. This is more of a customer or organization determined situation in which whether you would like just the totally qualitative test or an instrumented test, I think that will be available for this as well.

DR. KIM: Do you have any data for how much THC molecule is binding there?

DR. GOOD: We have. We have exposed the collector and done GC/MS pre and post. There is about 10 to 15% and it's a percentage basis really, binding to the collector. That's the most difficult one. For morphine and cocaine and most other things, you really don't see any significant binding.

DR. JONES (SAMI): As I read that slide on cocaine, you were seeing cocaine for up to two days after dosing?

DR. GOOD: Right, and this is from the methadone clinic trials in which they tell us when they took the drug and these are folks who have been urine positive and then oral fluid positive, so they are matched samples taken at the same time, one done by the certified laboratory using standard procedures. We went through this with opiates first. There was another product that really hasn't come out on the market. When we were looking at this, we just thought we would do a proof of principle to see how valid this was, and we were a little surprised about how far back we could see with this type of sample. These are probably significant users.

DR. JONES: The analyte was BE or cocaine?

DR. GOOD: There is cross reactivity in the antibody.

MR. ASCH: Sam, I like that new device, very exciting. What is the stability, projected stability and shelf life?

DR. NIEDBALA: Our specification is room temperature storage. We don't have the devices made long enough so I can tell you two or three years at room temp, but we have stressed as long as a month at 37, that is a pretty good standard for any immunoassay. We expect, like our alcohol test, we will have a minimum of a year and over time, that will push out to probably two years. The labels, which can be one of the weakest components, don't degrade at all.

DR. FOLEY: Sam, you had one slide where you were showing a collector. I lost track of how you move your sample from the collector to your reaction chamber.

DR. NIEDBALA: The collector, if you have seen our alcohol test, it's just a very simple plastic stick. This is the same stick with a different material on the end. That material, once the sample is collected, is simply delivered into a port. Again, the same basic idea as the alcohol test, except the port has a lock in mechanism. It is almost similar to some of the plastic devices you have probably seen for your kids when you are assembling toys. When it goes in, it snaps but doesn't come back out.

DR. FOLEY: As you move your sample from the mouth to the collector to the pad and to the device, what is your overall loss in THC?

DR. NIEDBALA: For THC, depending on the materials that we have tested, we could see losses as high as 50%. Generally, it's about 30 percent that can be lost.

MR. FORTUNA: If the person you are testing has used alcohol, does that interfere with the immunoassay?

DR. NIEDBALA: No. I can speak from the alcohol side of this that ten minutes is usually an adequate waiting period before you would collect a sample, and I also know we have tested people who have drank up to three drinks and then we have collected samples over time and not seen that to interfere.

DR. KADEHJIAN: For all the speakers on saliva testing, what is the availability of an appropriate saliva matrix for use in PT or quality control?

DR. GOOD: We have developed one internally and we will be providing that as a commercial product. It's not saliva per se, but most of your controls are not the real body fluid. They get modified for stability and shelf life and other things. This will be an appropriate fluid. A lot of it with oral fluid is you want something with the appropriate viscosity. If it's really thin like water, it isn't appropriate.

DR. NIEDBALA: Leo, I would also add just like for urine, someone a long time looked at all the constituents. That exists for saliva as well. You really can mimic it pretty nicely. I think it won't be too difficult. In an on-site test, I actually think it's going to be easier than what they are going to have to do for the lab based test, to tell you the truth, because there are different modes of collecting there. Here, it sounds like everybody is pretty much talking about direct sampling. For the guys who are going to do the PT constructs, I think they are going to have a somewhat easier job.

DR. SMITH: I'd like to make one more comment with regard to the relationship between alcohol testing and oral fluid or saliva testing. One of the things again in a Federal program is being able to standardize across the board the "reasonableness" of the search and seizure process, et cetera. As we have said before, with the exception of the alcohol testing, which is in fact a Fitness for Duty and does have a lot of tie to being under the influence or degradation in functioning, the urine test has never been so associated. I have noticed today with oral fluid that a lot of the cutoffs have been determined by the relationship to blood plasma, for example, or have much more of a hook, if you will, to under the influence. Would it be possible or do the manufacturers and distributors of these kinds of devices feel that eventually it may be possible to have truly equated measurements so that somebody who is subject to an NIDT urine based matrix and a person in another state or another part of the industry would be subject to a saliva screening test, could we be able to equate those in terms of detection and other issues?

DR. NIEDBALA: Let me comment on that. The only one who did that was Vina. I have actually seen that data before presented in Australia.

DR. SMITH: Was not the last presentation talking about cutoff levels at ten and 30 and those were linked more to blood plasma?

DR. GOOD: People have said that these levels do indicate under the influence.

DR. NIEDBALA: The numbers that are put up there on the slides I had were determined from clinical studies that have been conducted and have been part of the other working group that we have had for lab based testing. Those numbers have been looked at by another committee and there has been other DTAB meetings where we have brought those up. That's not new information and it is based on either dose studies in certain situations or as Carl has had, some random studies, but for at least cocaine, THC, opiates, there have been additional studies with saliva recently where a certain amount of drug was administered to humans and then monitored over periods of time and the urine and saliva looked at concurrently to look at windows of detection, and with two purposes.

One, we would all like to see what happens in those situations and what you can really detect, and second, because for purposes of FDA submissions. As a manufacturer, one of the questions we get is what is the minimum detectable dose you can see in a person.

There has been what have been referred to as prevalent studies as an example. For the laboratory based tests and system we are working on now, separate from today's meeting, we have gone out and tested 10,000 random subjects to look at what the hit rates are, what the concentrations in those groups are, potential interference that may appear from over the counter medications. There is a large body of data that even though this is technology that is next year, there is some basis for those cutoffs that you see.

In addition, if you look at something like THC, going down to 1 nanogram, and if you can do it accurately and competently, and again, people have brought up issues of manufacturers putting out substandard product, then you give yourself the widest windows of detection in certain situations. To what Carl had put up for THC as an example of a couple of hours, at 30 nanograms per mL, that's probably what you will see.

DR. SMITH: I think perhaps you are misunderstanding my question. It's not a question of how low you can go or to what extent you can extend your detection window. My only question had to do with whether we felt in terms of the technology that we could equate results so that they were the same as what we have.

DR. NIEDBALA: Impairment, detection windows or impairment? You don't have that in urine now. There are certain drugs in saliva that you can. That's in the literature. Pascal Kintz just published an article in the last few months which looked specifically at saliva and the correlation, and actually in JAT, there is one on codeine, as an example. It appears that for opiates, for cocaine and amphetamines, you probably have reasonable correlations to blood, but for things like THC, don't see it. Also depending on the route of administration, there is some variation. In urine, you are not going to get a correlation either.

DR. WILLETTE: Vina touched on very briefly that one of the problems in correlation with like amphetamines and basic drugs is the pH affects the saliva/plasma ratio because of protein binding.

DR. SMITH: My only point of bringing this up is as we look down the policy and program implementation road in a Federally mandated program across the board, would dealing with this particular alternative technology mean we might have to make some adjustments in how we have used such things as administrative cutoff levels, et cetera. That is my question.

DR. WILLETTE: All cutoffs established for any one of these matrices is an issue. The driving force has always been -- the limiting step has always been technology over the years. We reach an administrative point and the technology has continued to improve so we can actually detect at much lower levels with a lot of these. The cost factor comes in. There are a lot of factors that go into cutoffs. The original screening cutoffs were set by the limitations imposed on the manufacturers when they started developing immunoassays. It's a very valid question. It's a very complex answer.

MR. STEPHENSON: To go back to where I think Donna Smith was coming from, the issue is we have had a bifurcated program in the past in which we had impairment detection potential and consequences of a detection for alcohol, whereas we have never had that as an issue for urine based testing. Am I on target? Is that where you were going?

DR. SMITH: Right.

MR. STEPHENSON: The issue is looking down the road, now that this alternative technology presents a possibility of identifying current impairment or indicating that it could be interpreted as current impairment, what would be the implication for policy and program use if such a test were given. I think it is one of the issues that now is on the record and we will address that as part of the issues as we frame our policy development. It's one of the areas where for years, we have always had to say this is a demand reduction program, this is a deterrent program for urine based, it is not a current impairment detection program. Now you bring to that table that potential, so it is recognized.

MS. BOONE: I didn't mention reporting as a part of my slides. I just wanted to talk about the data management system that LabCorp offers. We have a software program that links the employer collector lab and MRO. The employer can schedule the event through Xnet corporate software, it walks the collector through all the collection processes, whether it's on-site or urine drug screen. It then notifies the lab that a test is coming and the lab can report that result to the MRO. There is also data management that exists from the collection site, demographics going to the employer MRO as well.

DR. CAPLAN: I just want to make one general comment since we are at the end. It was pointed out to me that there were a few occasions where some of the speakers might have kind of slammed the manufacturers on one issue or another about what they submit and how they deal with their things. I just wanted to set the record clear from our point of view.

We are very thankful that the manufacturers have provided the information they have today. It was highly useful. It was highly productive. Almost every manufacturer, and I've worked with lots of you directly, do have a true scientific element. There are obviously proprietary concerns. At least within drug testing, the science has been regularly shared. We have been on a number of other working groups where this has been clearly enunciated. I just wanted to make sure there is no perception that the manufacturers are not an integral part of this process and part of the scientific basis for which we are going to go forward and make these decisions. I apologize for any slamming, if anybody felt that might have occurred, but we do want to encounter all of the opinions and observations about the products. Thanks to the manufacturers for all their assistance and we know we will get that continuing in the future.

DR. BUSH: What a quality meeting this has been for us. Thank you for coming and being open with us. I'm going to be a little cynical here and say now you know how it feels to sit in the hot seat sometimes. If you took something personally that was said here, trust me, that happens every time somebody who has something important to say and then there is somebody else who has a different opinion and may voice that, you may feel like taking it personally, develop a thick skin.

Understand there are people who see this one step away from where you happen to be. Don't take it personally. This has been a wonderful quality meeting. I wish all our meetings were this good.

MR. STEPHENSON: I'd like to say our deep felt thanks and appreciation to our co-chairs of this meeting, Yale Caplan and Bob Willette. I think they have done a marvelous job of integrating their own individual research and background experiences into a very productive exchange and collection of information. I want to thank the DWP staff, Donna Bush, Walt Vogl, Charlie LoDico, and Richard Lipov who worked on this meeting.

[The meeting was concluded.]