about Epidemiology & the department

Epidemiology academic information

Epidemiology faculty

Epidemilogy resources

sites of interest to Epidemiology professionals

© 1999 

Last Updated

17 Aug 2001

Home HIV Screening Debate

XI International Conference on AIDS

Vancouver - July 8, 1996

Session Co-Chair: Helene Gayle, USA

Session Co-Chair: Margaret Fearon, Canada

Debater Affirmative: Ralph R. Frerichs, Department of Epidemiology (UCLA)

Debater Negative: Doris Schopper, Médecins sans Frontières, Geneva, Switzerland

Discussant: Michael Merson. Yale University School of Medicine, New Haven

Helene Gayle

Good afternoon, my name is Helene Gayle. I'm the Director of the National Center for HIV/STD and TB Prevention at the Centers for Disease Control and Prevention, and I'll be co-chairing this session with Margaret Fearon, who is a virologist for the Ottawa Department of Public Health. We're going to get started right away but first we wanted to explain the format of this debate so that people know what the ground rules are.

This debate has three participants and will have one participant presenting affirmative and another the negative position and then a discussion. Ralph Frerichs, who is the Chair and Professor of Epidemiology for the University of California at Los Angeles, will be giving an affirmative position. Doris Schopper, who is with Médecins sans Frontières, Geneva, Switzerland, will be giving the negative position, and Mike Merson, who is the Dean of the Department of Epidemiology and Public Health at the Yale University School of Medicine will be the discussant for this debate.

The two affirmative and negative speakers will talk ten minutes. They will then each have a five minute opportunity for rebuttal. Michael Merson will then be providing ten minutes of discussion, and then we'll be opening it up for about ten to twelve minutes for questions from the audience and then a wrap-up. We want to have this as interactive as possible but first of all give our presenters an opportunity to give their position. I'd like to just start by framing the discussion a little bit. There are a wide range of issues related to home or consumer access to HIV testing. There has been a lot of discussion about HIV home collection kits as well as the actual home test kits. What we'll be discussing here today are really issues related to home testing and not issues related to home collection. As many know, the home collection kit has already been licensed in the United States, is probably likely to be licensed many other places, but we felt that the issues that really need to be thought through are the issues that relate to the actual HIV home test kits and the ability to get test results in the home setting. That is really the area that needs debate, while recognizing that home collection kits for many parts of the world will never be a reality. So we wanted to make this discussion as broad as possible. I would like to turn it over to my co-chair for a moment who will also discuss briefly some of the issues as well as tell you what our overall resolution and framework for the debate will be.

Margaret Fearon

Thank you Helene. Although the issues around home testing in Canada are similar to those in the U.S., we have a slightly different perspective which I think is mainly due to the difference in our health care system. However, home tests will probably fall under the same regulations which govern laboratory-based testing, and therefore subject to the same rigorous laboratory evaluations which must be performed prior to approval for use in Canada. We are also grappling with the technical, ethical, psychosocial and public health issues that I'm sure you will hear about during the debate. Now I would like to read to you the resolution which the debaters will be discussing: "Resolved that HIV testing through home collection kits is an effective and appropriate addition to current testing and counseling services." I'd like to turn the floor over now to Ralph Ralph Frerichs.

Ralph Frerichs

Thank you very much. Could I have the first slide please? So the issue is, Why home HIV testing? The first reason is self-protection. We have two individuals who come together; one is infected and the other is uninfected, referred to here as discordant couples. They come in contact with the virus, then subsequently, whether it's a year later, two years later, five years later, likely both will become infected if the second one doesn't know that the first is infected. This process of transmitting the virus from one to the other has many years to take place. If we look at the natural history of the disease, there is a very long silent period extending possibly from nine to eleven years. I refer to these as the silent but deadly years when the transmission can take place if there is no knowledge that one is infected and the other is not.

The second reason for home-testing is self-caring. When we talk about self-caring for HIV infected persons, there are several areas to consider. The first is the early asymptomatic treatment of HIV. As you've heard in the plenary session that Scott Hammer gave on advances in retroviral therapy, there are many new drugs that are coming to the fore. Others have written and talked about this when they stated that new drugs need to hit early and hit hard. There's no way to hit early unless you know who is infected. So we need to have this knowledge in order to reduce viral load and make sure that the immune system is bolstered in its ability to respond to HIV.

The second point for self-caring is that HIV infected persons need to avoid contact with opportunistic infections. There are many infections out there. Preventing contact with these agents will only occur if the individual knows he or she is HIV-infected. And finally, for self-caring, HIV infected persons need to prevent transmission to their offsprings or loved ones. Without knowing of self-infection, this inadvertent transmission to those who are closed to them will often not occur.

The third reason for home testing is that people want them. A study was published last year but referred back to 1992, describing a survey done in the United States that asked, "Would you be likely to use home HIV tests, if they were available?" Fifty-two percent of those who thought that they had a high or medium chance of getting HIV/AIDS indicated "yes" that they would be very or somewhat likely to use home tests. The percentages decreased, but was still sizeable, among those in the low risk or no risk categories. What was startling to me is that this survey question was asked five years before any home test was even available in the United States, five years before there was any advertising or marketing of such a test, and yet already a high percentage of the population indicated that they would be very willing to use this sort of device.

So you may ask, "Why not rely on existing clinic-based HIV counseling and testing programs?" One problem is that such programs are not very effective at early detection. There is a startling study that was published by Wortley and colleagues last year that looked at over 2,000 individuals who had come down with AIDS and had been asked, "When did you first learn of your HIV infection?" It turns out that half first learned of their HIV infection within a year of coming down with AIDS symptoms. That meant that for eight to ten years they were running around with no notion that they were HIV-infected, thereby allowing the virus to quietly move from one individual to another.

The second reason why we can't completely rely on existing clinic-based programs is that even HIV-infected persons who come to clinics for testing are often not told of their laboratory findings. Several studies have been published, both in the United States and elsewhere, that show that individuals who are unknowingly HIV-infected when coming to HIV testing centers do not get their results. This non-reporting ranges from 17% among HIV infected teenagers in Houston all the way to 30% in OB women in a New York hospital. Overseas, 25% of HIV infected women coming to a clinic in Kenya didn't learn of their infection. Later in the same article, the authors reported that 65% of HIV-infected women were not notified of their infection, but told that they could get the results if they wanted them. Oddly enough, the authors made a virtue of not telling the Kenyan women that they were HIV-infected. Finally, you can see in this slide that among pregnant women in Rwanda, 35% of infected individuals didn't learn of their infection.

The next reason why we do not rely on existing clinic-based HIV counseling and testing programs is that they are too expensive for government funding of widespread early detection programs, unless there is prior personal screening. Here in this slide we are looking at a horizontal graph of the states in the United States that have government-funded testing programs. The lengthy green bar going from 0 to about 98 percent shows those many individuals who come to the clinic who are HIV-negative. Those little red sections on the ends of each horizontal bar are the ones that are HIV-positive. Only about two percent of individuals who come to government testing and counseling centers are HIV-positive. Money that is spent on the many HIV negative individuals cannot be spent on treatment for HIV positives. Money that is spent on HIV negative individuals cannot be spent on extended care for persons harboring the virus. So by supporting a system in which only two percent coming in are HIV-infected, we are spending an enormous amount of money on HIV-negative individuals that could be reallocated in more effective ways to slow down the epidemic.

The fourth reason why we should not rely on existing clinic-based HIV counseling and testing programs is poor coverage. Even in the United States, when asked in a national survey, "Have you ever had an HIV test for diagnostic reasons?", only about nine percent reported ever having such an HIV test. And it doesn't vary that much from state to state. In developing countries the situation is even worse. In many regions of the country there is no HIV testing for most people. Few clinics are set up in large cities, using the WHO or CDC style of counseling and testing activity. While such scattered clinics are often very well-run, they cover a very small number of individuals. Most people have no tests available, and thus falsely believe they are free of the virus.

The possible solution to our need for early detection is a simple, inexpensive HIV screening kit that people can use as often as they like in the privacy of the home. If they want to use it every three months, fine. If they want to use it every six months, fine, every year, fine. It's up to them. It should be sold in pharmacies or grocery stores similar to home pregnancy tests.

I went down to my local grocery store in Los Angeles, and noticed the many pregnancy tests that were on the shelves, shown here in the slide. Earlier in the United States there was much concern with such tests. Yet they are now very popular with consumers, and are widely available in stores throughout the United States.

In the future, I predict that we will have saliva tests available that I refer to as HIV indicators. Such tests are merely proposed, since they are not yet on store shelves. If the HIV indicator is negative, then concerned individuals are encouraged to again screen themselves in three to six months. If the HIV indicator is positive, based on two lines rather than one or maybe a color change, then the subjects would be encouraged to visit their physician and obtain further HIV testing and counseling. I refer to these screening devices as HIV indicators because they indicate the need for further testing. The results may be due to false reactions with other substances in saliva. People need to be warned about such false positive reactions. They also need to keep in mind that these are not diagnostic tests, they're screening tests.

I also feel that people in the future should screen themselves before coming to counseling and testing clinics for HIV confirmation. At-home HIV tests would then serve as the HIV screening indicators. Finally, the cost of these future home HIV tests should be kept low with subsidized prices similar to what we now have with condoms. That means that governments and NGOs need to get involved, so that tests like this become more widely available to those who can benefit most from them. Thank you very much.

Margaret Fearon

Doris Schopper will now give the negative position.

Doris Schopper

As Helene Gail mentioned in her introductory remarks, the term "HIV home-testing" creates confusion as it is used to refer to two different things. HIV home-testing is immediately associated with do-it-yourself tests. That's what Ralph Frerichs just talked about. But we should clearly distinguish between the home collection tests, which implies that the analysis of the specimen is carried out in a laboratory and the tests results are provided by a third person, and the home self-test that can be carried out entirely at home without involvement of a third party.

When necessary I will distinguish between these two types of home tests. Whenever I refer to both systems simultaneously, I will use the term, "home testing." First slide please. As Ralph Frerichs just mentioned, the central argument for HIV home testing is that it would considerably increase the number of people who have access to and may use an HIV test.

However, to my knowledge, there's only one study which provides some insight into the potential for home collection HIV testing. Twenty-nine percent of adults interviewed during a national household survey in the United States said they were likely to use a home collection device for testing, and twenty-two percent said they would prefer this method to clinic-based voluntary counseling and testing. Persons with less education, with lower incomes of non-Caucasian origin, and who did not know where to go for an HIV test were more likely to say they would use home collection testing than the other groups. However, respondents were not informed about the probable price to pay, estimated at 30 or 40 U.S. dollars, and were not asked about the reasons for preferring home collection testing. In addition, no questions were asked about the intended timing of testing, in particular, in relation to past or present risk behavior. There are absolutely no similar data on the potential demand for home self-testing.

Let's now look at some of the technical problems of home testing. First of all, what is the accuracy as compared to the traditional HIV tests that use serum or plasma? The main technical problem is that the antibody levels in saliva and urine are considerably lower as compared to those in serum or plasma. Several studies have been conducted comparing the results obtained with saliva and urine to those of serum and different collection devices have been evaluated. You can see here that the reported sensitivities and specificities for oral fluids, urine, and capillary blood, are lower than for serum. This, and the difficulty to confirm a positive result on non-serum plasma specimens, is of concern when these tests are used for diagnostic purposes. In addition, most of these results are for tests which can only be done in a laboratory, and few data are currently available for essays which can be potentially used as home self-tests. Finally, data presented here were obtained under optimal conditions by trained technicians. It may not reflect a real-life situation as simple rapid tests that request subjective interpretation can easily be misinterpreted by laypersons.

Beyond this issue, there are a number of technical problems with home-testing, as shown in this slide. The quality of the test at the time of use depends not only on the manufacturer, but also on transport and storage at the sales point and at home. The quality of the sample collected depends on the complexity or ease of the sampling method and the understanding of the user. An additional problem is the correct understanding of the window period as the incentive may be to test rapidly after risk of exposure. More generally, the level of education needed to correctly understand the testing procedure should be of concern. For home self-tests, there are additional issues such as errors of manipulation if done by laypersons, interpretation of results, and action to be taken for confirmatory testing. I would like to point again to the fact that the accuracy of the new generation of capillary blood, urine and saliva tests which could potentially be used as home self-tests, has not yet been evaluated extensively, and their reliability if performed by laypersons has not been established.

Beyond these technical issues, there are two other major questions with regard to the risk of home testing. Does testing without pre-test and post-test counseling present risk to the individual tested? And is there greater risk of systematic abuse of HIV testing by third parties than with laboratory-based testing? In order to guarantee the individual benefits from an HIV test, to reduce the fear and negative effects of testing, and to increase the possible public health benefits, counseling has been promoted as an essential element of voluntary testing. However, it has been shown that in many countries, testing is done without proper counseling, or counseling occurs only after the test results are known. Proponents of home testing thus argue that pre-test counseling is not essential and that counseling after home collection or self-testing could be provided more efficiently with telephone contact which is cheaper and would guarantee anonymity. This scenario, though, relies on access to a telephone communications system, and some type of quality control of the counseling process.

In addition, it is not clear who would provide the service and pay for it. If this is the responsibility of the test manufacturer, as has been suggested, the incentive to train, supervise, and maintain qualified counselors is limited if this is not linked to a periodic accreditation system. Anonymity will only be guaranteed until the HIV-positive person goes to a health service, and in most developing countries, telephone counseling would of course be highly impractical for the majority of the population. Home testing may thus lead, on the one hand, to an increased number of false positives, as positive results are taken at face value after the first home test. This might be a problem, especially in low-prevalence populations. On the other hand, an increased number of false negatives may occur due to testing shortly after risk behavior without retesting after three months. This should be a particular concern in areas of high incidence of HIV infection.

One of the major concerns voiced by opponents of home testing is the potential abuse by institutions, employers, insurance companies, police, border controls, and by sexual partners, in particularly, men against women. Proponents have argued that possibilities for abuse may not be very different from those already existing, and that an otherwise beneficial product should not be outlawed because it may be potentially abused. There are at least two reasons why home testing may be more amenable to abuse than laboratory-based testing. It is easier to do under coercion and there is less guarantee of confidentiality as the result is directly available in the home or by a telephone call. Home self-tests, which would in addition provide immediate results and could be used directly at border controls, by future employers and by sexual partners, with little or no consent of the person tested. Given the level of persistent discrimination and stigmatization in many countries, this potential for abuse is worrisome.

Overall, the risks and potential negative consequences for home tests should be expected to be much greater in situations where the powerless are not well-protected by law or regulations, where the status of women is not equal to that of men, where quality control of medical devices and procedures is difficult, where regulations are not enforced, where literacy rates are low, and where health services are either scanty or difficult to access.

I think we should realize that the introduction of HIV home tests on the market is nothing less than a revolution in the history of health and medicine. This would be the first time ever that an individual has the ability to diagnose, by himself, a chronic, incurable, and most probably fatal condition. The only other home self-test to detect a health condition that was ever used widely in developed countries is the pregnancy test. Although the consequences of a positive pregnancy test are quite different from those of a positive HIV test, I would like to briefly review some of the lessons learned from the home-use of pregnancy tests that have been available now for almost twenty years. To my surprise, the scientific literature on home use of pregnancy tests is extremely limited and I could find no information for developing countries. Over-the-counter pregnancy tests became available in 1976. The FDA, after reviewing data on several of these tests, stated that they were reasonably accurate for preliminary checkups. The arguments used against making home pregnancy tests widely available were inexperience of lay people in testing procedures, difficulty to evaluate false positive or false negative results, and that a pelvic examination should supplement the tests. Arguments in favor of home testing included their lower cost, desire of anonymity, and possibly earlier diagnosis leading to better prenatal care.

All of these arguments remind us singularly of the current debate on home HIV tests. Although there seem to be many questions, no assessment of the use of pregnancy tests at home was apparently done for a number of years. It was only in 1982 that a U.S.-based study found that in a sample of 144 pregnant women, the overall false negative rate of the home pregnancy test was 24%, and total compliance with test instructions was only 32%. Test accuracy was directly related to the educational level of the user. In 1986, another study showed that accuracy of test results at home ranged from 46 to 89%, differing from the 97.4% claimed by the manufacturers.

Further issues identified in these two and other studies was failure to respect the nine-day waiting period after menses and failure to repeat an initially negative test. As with HIV testing, the lower cost of self-testing has been one argument in favor of home pregnancy tests, which has, however, not been well-documented. In addition, the better access to prenatal care and positive behavior change has never been documented. Lessons thus learned is that test results will on average be less accurate when test is done by a layperson and that the user needs to be well-educated to understand operating procedures.

Thus, in conclusion, I would say that it would be highly irresponsible to provide these new tests to laypersons for self-testing before answering some of the most essential technical, psychological and social questions, including the following: What is the intrinsic accuracy of the currently available HIV home-collection, home self-test? What is their reliability if used by laypersons, including the timing after risk behavior? What is the probability that persons identifying themselves as HIV-positive will not seek confirmatory testing, and what would be the consequences? How cost-effective are these tests? What is the demand for and accessibility of these tests in different environments? What are the express needs and fears of users? What is the importance of providing counseling before and after the tests? And how can post-test counseling be provided in low-resource settings? Finally, what action will individuals take after receiving an HIV-test result at home? And what is the potential for abuse in different settings and how can it be prevented?

All these questions relate to both home collection and in-home testing. In addition, we should ensure that some minimum requirements be fulfilled for any test before approval, including internal control mechanism after marketing which validates the test results. No test should be marketed in another country before having been approved by the regulatory body of the country of production, as there's a real danger that tests of low quality are brought on the market in developing countries, and clear guidelines must be provided with the test on how to confirm a positive result.

I guess we would all agree that people should have access to information that enables them to control their lives. But at the same time, we need to acknowledge that external conditions can make the same technology favorable or dangerous. If we want to abide by the Hippocratic principle of first do no harm, we must ensure that the benefits of a new technology outweigh its negative consequences. It is thus essential that sound research on the questions raised be undertaken immediately in a variety of settings, including developing countries, and that decisions on the use of home-collection, home self-tests by laypersons be based on the results.

Margaret Fearon

Ralph Frerichs, you now have five minutes for the rebuttal of the negative position.

Ralph Frerichs

I hesitate to comment about the pregnancy test. I have two children, but we didn't use it in diagnosing either one, so I don't know that much about them. There are a number of other screening tests in use in the United States -- and again, I look at home HIV tests as screening tests. Most women know about breast self-examinations. When doing such personal screening, there is no immediate counseling to help interpret the findings. Pap smears are routinely done, although in a medical setting, but again it's a screening test and not a perfect test. It does not have 100% sensitivity and specificity but rather is a test that kicks you into the system for confirmation. We now have home blood cholesterol tests that people buy in pharmacies and use without counseling. If you get a high blood cholesterol reading it does not mean that you're going to have a heart attack tomorrow. It indicates though, that you need to go and consult with your physician and address this problem. We have screening tests like the PSA test for prostatic cancer that do not have inordinately high sensitivity and specificity because they're screening tests -- they indicate the need for more testing. So this issue of perfection is something that is always going to be elusive if you demand that screening tests have the same quality as diagnostic test. They intrinsically cannot have that, although saliva tests for HIV are nearly indistinguishable from blood tests.

As I have already pointed out, when we look at the sensitivity and specificity of the test, we need to move beyond that if we're going to take a public health view of the problem. Rather than focusing excessively on the technical aspects of the test, we need to consider the sensitivity and specificity of delivering accurate information to those who are HIV-infected (that is, they are HIV positive) and to those who are free of the disease (namely that they are HIV negative). When we look at existing testing and counseling programs, they are so inefficient that many HIV-infected people who come to them do not get their results. This inefficient situation is what we have to compare to when talking about a screening program. We can't focus only on the sensitivity and specificity of the test.

Ralph Schopper and I agree that people should have access to information that enables them to control their lives. This is probably the central point where I think we agree. We also agree that the changes we are talking about should not come about in a dramatic way, but rather ought to be thought about and dealt with in experimental settings, in which we look at these things more carefully. Where we seem to differ is that I'm saying it's time to move ahead. I think that we need to push industry to develop these home HIV tests, we need to be willing to go out and experiment with their use, going out and setting up different trials and study sites, and seeing exactly what happens. Finally, we can't let the fear of giving somebody a test or a screening device without counseling, interfere with moving ahead and seeing whether these mental barriers we have created are indeed real.

So I'm suggesting at this point that we move forward, that we institute some studies, that we encourage manufacturers to develop these tests, and that we be willing to try them to see whether they create the benefits that many of us think will occur. Thank you.

Margaret Fearon

Dr. Schopper, you have five minutes for the rebuttal of the affirmative position.

Dr. Schopper

As Dr. Frerichs said, we believe that people should have access to information. Where we strongly disagree is what the effect of the home test should be. Dr. Frerichs' main argument is around self- protection, self-caring, and that people want them. I think I already answered in my presentation that we don't have any good data on what people want them and what people would do with them.

Secondly, with respect to the issues of self-protection and self-caring, he says they will do better than current voluntary counseling and testing. I think that is totally unproven. We should remind ourselves that when it comes to protective behavior after voluntary counseling and testing, that we have very mixed evidence, and that the only clear evidence we have is that discordant couples who are tested and counseled together may have positive behavior change. In heterosexuals or homosexuals who are tested alone, the evidence is very mixed, and we have several studies which show that HIV-negatives may engage in higher-risk behavior after having their test results. Now I leave it to you to think what happens when you do the test alone in your home. Will you be better off with regard to behavior change than if you had at least engaged in a discussion with somebody who has given you information and discussed personal, individual choices for behavior change.

I would also like to say that there are some flawed assumptions in what Dr. Frerichs says about tests leading to change in behavior. We have many studies which show behavior change without testing. So testing can only be one small addition to our whole attitude towards the HIV epidemic, as we know. He also assumes that there's no benefit of voluntary counseling and testing to those who are negative, which I would also contest, as it's probably very important for those who are negative but have engaged in risk behaviors to be able to have this personal interchange. I again want to point to the fact that we're talking here with a very biased view reflective of the U.S. and Canada. In developing countries, as I pointed out, many of the issues are absolutely not clarified. Home HIV tests could be an extremely dangerous technology. When I see Dr. Frerichs' slide which shows this huge bulk of people in developing countries who don't have access to testing, I would very much question that it is those people who will have access to home tests and who will be able to use home tests in a coherent way, given the lack of education, lack of literacy, lack of confidentiality in the home and so forth.

So I would argue that poor coverage will not be achieved by home tests, but that what we should do is try to improve our voluntary counseling and testing strategies. And I think that is where this new technology can have a great role to play. I know from some initial studies which were presented at the Kampala conference -- I don't think they are presented here -- which showed that when you use rapid tests in the counseling and testing setting, that people actually get access to their test result right away. Pointing to the studies Dr. Frerichs mentioned where he said people did not get their test results, in several of these studies it's not that they didn't get their test results, it's that they did not come back to get their test results, which is a very different issue. Why do people go away and don't want their test results afterwards? Well perhaps it is those people one could help with their test results if they get their test results right away. I think we need to think much more about current barriers to voluntary counseling and testing and what makes it difficult for people in those settings and why do people not have access to voluntary counseling settings instead of dumping the home-testing issue on people. Offering such home tests may also lead to government's bailing out on providing voluntary counseling and testing services.

Margaret Fearon

Dr. Merson now has ten minutes for discussion.

Michael Merson

Well you've heard arguments for and against home testing. I've assumed that my task is to use the best of these arguments to suggest a course we should follow. My remarks are primarily based on a discussion that took place yesterday in a pre-conference workshop on home self-testing supported by the Henry J. Kaiser Family Foundation, attended by 35 experts from a wide range of disciplines and countries, including industry representatives.

I do not mean to suggest that all those participating in that meeting would agree with my views -- this was not a consensus conference -- but I would hope that most of them would feel that they are hearing now a reasonable reflection of the main conclusions reached at that workshop.

Before stating these conclusions, I want to make three comments to place the issue of home self-testing in perspective. First, the decision as to whether and how to proceed with any intervention necessitates weighing the potential serious risk of this intervention against its individual and public health benefits. I can assure you this task of assigning risk and benefits is not simple for home self-testing for HIV infection. As Doris has said, if such tests are available it would mean for the first time that a layperson would be able to diagnose, on his or her own, a life-threatening illness. Second, often in the practice of public health, we need to make decisions without all the information we would like to have. The urgency of the AIDS epidemic has placed us in this situation time after time during the last ten years. We should try to avoid this here. But the fact that there are a number of home tests already available on the grey market will not make this easy. And third, we need to be ready to change our views on any intervention as more information about its risks and benefits become available. This is always true, but as we have learned in the AIDS epidemic, it's particularly the case when one deals with a new disease.

So let's start with a few facts. The information we have, and it's mostly from the United States, indicates that today half of those infected, including many who practice high-risk behavior, are unaware of their HIV status, and half of those who test positive have their first test within a year of developing an AIDS-related illness. I suspect this situation is true elsewhere. We also know, and you've seen some data from Ralph, that some one-third of those tested do not return to receive their results for fear of loss of confidentiality, stigma, and failure to cope with a positive diagnosis. This is not surprising, since in many testing sites, counseling services are inadequate or non-existent. It is these fears which makes the idea of an anonymous home test appealing. In fact, there is evidence that you heard from one study in the United States, that the availability of such a test would at least double the number of those tested, particularly poor, young persons of color who distrust the traditional health care system and are at higher risk of infection.

A home test could thus, in principle, increase access to HIV testing, and assure anonymity and confidentiality. Why is this important? Well first, it could simply make it possible for more people to know their status, ranging from the "worried well" to the "not-worried, not well".

Second, it could allow more HIV-infected persons to receive antiretroviral therapy early in the course of illness. Early therapy now appears to decrease the rate of disease progression and thus increase life expectancy. The more effective our therapy becomes, the more important this will be.

Thirdly, it could result in a decrease in mother-to-child transmission by allowing more child- bearing women to know their diagnosis. This will help them decide if they want to have children, and if they are pregnant, to receive antiretroviral therapy. Of course, to benefit in these two ways from antiretroviral therapy, those infected must have access to these drugs. We are regrettably a long way from achieving this situation, especially among poorer and underserved populations the world over.

A fourth potential benefit of home testing that you've heard is that it should decrease the rate of HIV transmission and thus reduce the number of new infections. One would hope that the more people know their status, the less likely the virus would be transmitted. Unfortunately, as you also have heard, except in discordant couples where both partners have received counseling, we do not have evidence as yet that HIV testing and counseling brings about behavior change. Perhaps we will in the future.

Another benefit of home testing is that it would allow some of the large amounts of money now spent for on-site testing and counseling programs to be used for other AIDS prevention and care activities.

Well, if you accept these as justification for moving forward with a home test, the next question is to ask is, What steps are required? What questions must be answered for all of us to feel comfortable with its use? I see at least three major concerns.

First, we need to have a test that in a consumer's hand, is as accurate as today's standard blood test and as simple to perform and read. While it is possible that we may be able eventually to develop a test that will make a true self-diagnosis, as you have also heard, it is likely that the first home test that could be available will be screening tests, using blood or saliva. They are likely to have a higher sensitivity than specificity, and thus to result in a number of false positive reactions. This will mean that persons who test positive will need to obtain a confirmatory test from the health care system. No doubt this has its advantages, as it provides better opportunities for counseling and care. Its disadvantage is that it compels the positive reactor to live with a diagnosis that is sometimes not definitive.

Secondly, we need to know much more about how persons would respond to the test result in the absence of a trained, face-to-face or telephone counselor. Emotions and thoughts such as fear, anger, depression and hopelessness, if the test is positive, and relief and invulnerability if the test is negative, can lead to a myriad of behavior responses. We will need to find out more about the severity and determinants of these responses and their consequences in populations that will make use of home tests, and what could be done to ameliorate them.

Thirdly, we need to take necessary steps to ensure that these tests, on the one hand, are not abused for purposes of employment, housing, travel or insurance, and on the other hand, do not result in physical or sexual abuse of women who will be often blamed for infecting their male partners. This worries me the most, as there are many countries where laws and regulations are not in place, or if they are in place, not enforced, that protect persons from being coerced to test, or from the physical violence that results from a positive test.

There are a few other concerns as well. Some of us will need to give up our desire for mandatory reporting of HIV positive cases, which in some places could make it more difficult to track the epidemic. I view this as a minor issue, especially since there are other means to do this. More importantly, we need to be sure that the availability of such tests does not undermine the network of community clinics, private and government-run, that presently offer HIV testing and counseling. There will be persons who will prefer to be tested in such settings, and they must have the opportunity to do so. It would be worth considering the providing of home test kits at reduced price at these clinics for those who would wish to use them at home.

Lastly, we would need to be sure that such test kits are properly labeled and advertised. Some issues to think about. How would it be best to ensure that the user of a home screening test obtains a confirmatory test, if the home test is positive? How will we make the consumer and public understand that a positive test means the presence of antibody and not virus in the saliva? Can we find an innovative means of helping consumers and perhaps their partners seek and obtain counseling if their test is negative, to discuss safer sex practices? As long as antiretroviral drugs are available, it is appropriate to market home self-tests on the basis of their value in obtaining early treatment. What additional evidence do we need to also market these tests as a tool for prevention? Is it ethical to promote and sell the product -- this is Doris' point -- in areas such as much of Africa, for instance, with low literacy rates, scanty health services, and inadequate quality control procedures? And last, can the cost of these tests be made affordable to those at the greatest risk of infection?

Now for sure, we will want to examine carefully the initial experience with home access tests to see what lessons they provide for self-testing. Information on who uses these tests and why, and their experiences with telephone counseling, will be particularly useful. One question to ask is, will they result in more people getting tested, or will persons using home self-tests be those persons who would have anyway used... .

The availability of a home test will raise one other important issue I want to mention in closing, and that is its social impact. I would like to think that placing the decision to be tested in an individual's hand just might remove the stigma surrounding HIV infection, just might allow people to discuss their disease openly without fear. If this one fact turned out to be true, it might be reason alone to promote home self-testing.

In conclusion, let me say that I think we should cautiously -- I repeat, cautiously -- and responsibly -- I repeat, responsibly -- pursue a research and development agenda for a home test for HIV infection. This means refraining from moralistic arguments. It means resisting temptations to make a quick profit. It means moving forward rapidly with a rational research agenda, to answer the questions I and others have raised. In the end, let us base our policy decisions not on the availability of the technology, but on the sound scientific data and consultation with consumers, including persons living with HIV and AIDS. Above all, the science, and not the technology, should drive us. Thank you.

Helene Gayle

I think that's been an excellent presentation of pros and cons and Dr. Merson's final discussion pulling the issue together. We'd like to open it up now for about ten minutes of questions you could ask any of our panelists here. We have two microphones in each of the aisles, so if people could just come to the mikes.

First Question -- person not identified

As a male that's been living with HIV for fourteen years, I cannot imagine testing myself. I can't. I mean, you (Dr. Frerichs) likened it to a woman doing a breast check, but that woman went into her doctor, was shown how to do it, was told all the consequences of it, it could be benign, it could be malignant, she was counseled. Now to me, counseling is what has kept me alive, the positive attitude, and the help and support. But to throw it out and say here, test yourself, I think we're really asking for trouble. Thank you.

Second Question - Neil Constantine, University of Maryland

Just a comment on the technical concerns that were raised. Dr. Schopper showed some data about the sensitivity and specificity of saliva tests. Those were a bit misleading. If you looked at all the test studies that were done, they may be correct, but if you single out and start to sort out the studies that use proper collection devices and assays that were designed specifically for testing oral fluids, the data are much better than that. In fact, and Dr. Frerichs has done a number of these studies, they're very close to a hundred percent all around. We're reporting on Thursday at the poster session on the use of two rapid assays and their sensitivity and specificity are essentially 100%. So the technology is there, I don't think there should be too much question that the assays now can detect antibodies in oral fluid. However I do also recognize that our studies and others were done under ideal conditions, as was mentioned, and that we do have to be concerned about the conditions at home when people use these tests. Thank you.

Third Question -- person not identified

I would like to discuss economical arguments which have been raised by the first speaker (Dr. Frerichs). You said that the high number of HIV negative persons implies a high cost for society which could be allocated to other care, so I would like to discuss these points. First, you did not include in the calculation the benefits of counseling, and I mean, we have to consider that. Second point. Because of the availability and easiness of the doing this test, it is very likely that home HIV tests will be used very widely and that people will repeat the test very often. So I'm very skeptical of this technical and economical arguments. And last, in most countries when these tests are self- prescribed they are not reimbursed by insurance -- that means that probably poor people will not have access to this test if they aren't prescribed. So, it was just these three points I wanted to raise. Thank you.

Helen Gayle

Thank you. And may I remind people, although the comments are good, we'd also like to take any questions that people might have for the presenters. Next mike.

Fourth Question -- person not identified

I'll try to contrive my comments into a question. I'm a medical case worker and I work specifically with Asians and Pacific Islanders in San Francisco. Most of the patients that I'm serving come in and receive services when their CD4 is less than 100 with about half of those actually presenting at CD4 less than 50. Yet the main issue that I see patients having to deal with are mental health concerns. The level of anxiety and denial that people deal with I think goes beyond almost all other medical concerns initially around HIV, and given the fact that you're dealing with a late-stage or advanced-stage population that certainly should be looking at PCP prophylaxis as their primary concern. I want to comment on just a few things.

It's my understanding from some CDC data that if you look at youth who present at STD clinics, 29% of those youth will present with more than one STD. One of my concerns in terms of the home testing kits is that these are certainly people who are engaging in what would be risky behavior and what we're doing is potentially separating patients from clinicians for appropriate screening, not only for HIV, but for all other STDs which these populations are at risk of. And finally, I would like to say that I think that we cannot underestimate the risk of coercion to the women who are in relationships or outside of relationships and whose partners will use this and manipulate them.

Fifth Question -- Elizabeth Dax from Australia

We've heard in this discussion several references to sensitivity and specificity. I would like to point out that they're probably not the relevant issues here, that we ought to be talking about predictive values. And even in the blood-care setting, when we're using highly specific and sensitive tests, in fact the predictive values of the tests are extremely low, even in low-prevalence populations. And so that if you have a test of 99.9% sensitivity in a population where the prevalence is 1%, your predictive value is something less than 30%. Now I'd like to address the question please to Dr. Frerichs. If one in three people who are testing themselves are going to have a reactive test, and you've already shown that many people do not come back for counseling or follow up, how do we put that in that context please?

Response by Ralph Frerichs

When talking about the predictive value of a positive test -- different from the sensitivity and specificity -- you have to take into account the prevalence within the population. Maybe the best way for me to frame this is to consider the alternatives to home testing. If somebody decides that he or she wants to go to a HIV counseling and testing center for a diagnosis, there's self-screening that occurs prior to getting the test. They may sense that they're at risk based on health education and knowledge of personal behavior, so they come forward. Or it may be that a friend or lover who knows of their risky behavior suggests that they go in for a test. The sensitivity and specificity of the self-screening or friend-screening process, the sensitivity and specificity of looking somebody in the eye and saying you may or may not be HIV-positive and encouraging them to go in for testing, is not very high. Each of these are forms of screening. What you're talking about with home HIV tests, is screening that is reasonably good and certainly much better than the "I sense my risk" or "look them in the eye" alternatives. I happen to agree with Dr. Constantine, that these tests in many settings have been very, very good. The sensitivity and specificity are so high that they compare easily with any other screening tests that we have out there and are much better than most. But invariably, if you're going to have a low-risk population, a very low-risk population, you're always going to find a situation where you may have 20, 30 or 40% predictive value. But that compares in the United States with the 1 to 2% predictive value that you now have when bringing people into testing centers (low specificity of system) and the 40-50% of unrecognized HIV infected persons (low sensitivity of system). So it's the concentration process of screening that you should focus on, bringing more individuals who are HIV positive to testing and counseling centers, thereby setting up a more efficient operation so that you can spend more money on alternative ways to deal with the problems that impact HIV-infected people.

Helene Gayle

Because we're running out of time, we'll take the last four folks who are already lined up and we'd ask you to just keep your comments very brief or your questions, if you have questions, very brief.

Sixth Question -- Mary Jane Rotheram, UCLA

I was actually a little surprised that you didn't add to your arguments that people who find out they're HIV positive make changes. That in fact 80% disclose to their partners that they're positive, that condom use goes up dramatically to a mean of about 77%, and that about 35 to 66%, depending on the population, use condoms 100% of the time. That's with women with AIDS, with HIV-positive adolescents, and with men who have sex with men. And I think those are the kind of data that, if in fact we could move that early detection threshold down, would be very powerful in terms of changing some of the dynamics of HIV transmission.

Seventh Question -- Harry Nichols, physician

I understand something of the technicalities and the clinical implications of these things. We have already heard it said during this conference that we've got to listen to the people concerned, and therefore I thank you very much, the first speaker from this microphone, for his comments, and I am very aware of the part that fear can play. I will be responsible when I get back home for advising some of my colleagues in the licensing part of our government and in making certain public health policy recommendations, so I have a great desire to listen more to what some of the people who are going to use it have got to say. And finally my question is to Dr. Schopper, please may I have a copy of your ten questions before I leave here?

Eighth Question -- Steve Wakefield, Night Ministry

We provide shelter to the homeless and runaway youth and health care on the streets for sex workers at night. I'm clear that counseling hasn't taken away the stigma for HIV/AIDS, and that whether we debate it, whether the technology and science exists today for home test kits, it will in a few short months, weeks, years, no matter what we want to do about it. And I guess my question is for the last speaker (Dr. Merson), and I'm interested in who would do the research that we're talking about and how we think the research would answer questions, particularly with high-risk young men who don't trust health care systems, who don't go to places where the testing and counseling currently exists, and who, a large number of them if they test once HIV-negative, are going to still continue risky behaviors and might benefit from having a test kit available to them at some point. I mean, I just don't understand who's going to design this research. I also stand here as an HIV-negative man who's lived with the stigma of HIV testing, and the risk of being positive, and been tested every six months for the last fourteen years, and I wonder every time what that means if I go in for a test, what that means for my insurance, what that means with my employer, and you know, just where's the real-life lab going to exist and who's going to design the test? Who are we suggesting could do the research necessary that's going to get us the answers?

Response by Michael Merson

The feeling of the meeting we had is that we can't just leave the situation as it is. We must address some of the issues that home testing raises in a research and development agenda. I believe that there are certainly research groups in the United States and around the world, including research groups that have a lot of experience with outreach and working with people with HIV and AIDS that can deal with some of the questions that have been raised here. I don't want to get into a debate. I don't think it should be funded by industry per se, but I think there are bodies that ought to fund this kind of research in the public sector to get some of these answers. The only point I would say is that -- and I want to be clear -- this is not a question of not counseling. No one says that people who are infected should not be counseled. The issue here is the coupling or decoupling of the initial testing with the counseling. One of the research questions that would need to be addressed -- the critical one -- is what happens to people when they get a diagnosis if they don't have immediate access to a counselor. I think that question is a reasonable one to ask. I'm not saying that the results will be positive or negative, but participants in our Kaiser Foundation sponsored meeting felt we needed to learn more about how people react when they get a diagnosis. There's very little information on that. I know a lot of you have personal experience with this, but there is not a lot to go by in the published literature.

Ninth Question -- Tim Sankary, UCLA

I'm a Fogarty fellow assigned to Japan, but am going to speak from past experience. As a physician, I have personally tested over 5,000 people and counseled them for the HIV test so I'm very aware of the kinds of reactions we can have. But I would like to address my remarks to my current work where I think this test would potentially have an application -- namely, with marginalized segments of society. I work with foreign female prostitutes in Japan. Among these women there is a high level of interest in a test that they could use on their own. Access to testing, while widely available in Japan, is difficult and threatening for foreigners, especially since the explanatory material is all in Japanese. So in this kind of setting there is high interest and potential use among marginalized segments of society, many of whom fear establishment kinds of clinic and anonymous test sites. For them, home testing is a means of empowerment. I would challenge the opponents of such home testing to reflect more on what they mean when they say they allow individuals to diagnose their infection. Rather than a diagnostic test, we're calling it a screening test. Just as with a self-breast exam, if you find a lump it's very scary but it prompts you to go in and seek the medical care. In the same way as a self-testicular exam would prompt you into care. And I disagree that everyone doing a breast exam has had the counseling, I think they can learn it from a pamphlet. As a screening test, the home HIV test has a very valuable aspect to it. Thank you.

Helene Gayle

You want to wrap up?

Margaret Fearon

I'd just like to very quickly wrap up by thanking our participants for a very clear presentation of all of the issues that are involved with home testing, and I think it's obvious it's a very controversial issue and one in which there's a lot of disagreement. But I think there's one thing that we would all agree on is that, whatever testing we use, whether it be home testing or laboratory-based testing, that we have to ensure that it's the highest quality test that's available and that it has a reasonable cost to encourage use. And I think we should also ensure that quality is not compromised for the sake of convenience or for the sake of profit-making. Thank you very much.