14

In general, I believe that the extensive medical histories and physical and laboratory examinations carried out on workers and prospective workers too often lack a validated basis or adequate relevance to the particular exposures of the worker's environment. As you know, Rockey, Fantel, and I recently published a very critical analysis of the use of pre-employment back x-rays; Dr. Rockey is testifying before the Committee today, as are representatives of the railroad industry, where the x-ray screen is still widely utilized, despite its poor predictive value. There is a genetic twist on the screening for low-back problems. The histocompatibility antigen B-27 is strongly associated with one type of low back problem, called ankylosing spondylitis; some physicians have recommended doing this test, which involves no radiation exposure. Even though individuals with B-27 have a roughly 200

times higher risk of ankylosing spondylitis than do people without B-27, the absolute risk is still low (about 5 percent for symptoms that are clear-cut, perhaps up to 20 percent for suggestive symptoms). Not only would a positive B-27 test not be basis for excluding workers, but also this lab test would miss the rare patient with some remediable back condition!

With regard to specific genetic markers, we know a great deal about genetic variation for certain traits in the general population. Unfortunately, research into the consequences, if any, of such variation in the workplace settings is only beginning. We can suggest here the criteria which should guide the development of any research results into potential tests:

the prevalence of the predisposing trait must be sufficient to be detected in usual worker populations; usually this will mean at least 5 percent prevalence but "sufficient" will depend upon the reliability and cost of testing and the relative risk and its consequences

O the prevalence in various subgroups in the population should be determined, since genetic traits often occur with differing prevalence in different ethnic or geographic groups

15

the relative risk for a significant adverse outcome must be at least 3 (risk for those with the trait compared with those lacking the trait), based upon experience with relative risks for histocompatibility types associated with various diseases; preferably the relative risk should exceed 10 times, so as to enhance the likelihood that the particular tested predisposition is, indeed, putting those workers at significantly higher risks than other presumed but non-testable predispositions within the overall population

the test itself must be reliable, inexpensive, and well-characterized

for false-positives and false-negatives

O there must be a clear understanding with management and workers about how test results are to be utilized, what actions might be based upon the results. There is no point in using unvalidated tests for screening purposes. Also, there is no point in assessing genetic variants in situations with overall low risk. For example, the one remaining possible hazard for sickle-cell trait is exposure to low-oxygen enviroments in flight accidents or high-altitude military or mountainclimbing exercises. The Department of Defense as of January, 1981, finally recognized that there was no factual basis to continue excluding blacks with sickle-cell rrait from the Air Force Academy or from pilot training; some had slipped through the screens anyway and preliminary follow-ups are benign. However, with the new policy of non-exclusion the military is launching careful surveillance studies to monitor and assess this question with data.

If and when properly developed and validated tests are available, the usual processes for determining standards and recommending tests should apply. NIOSH, as recently as a presentation yesterday to a National Academy Panel on Risk Assessment, has stressed its interest in identifying and protecting especially susceptible workers.

It is premature to speculate what should be done to the workplace or with the worker practices or both in response to a particular, yet to be discovered predisposing condition. Efficient, ethical, and effective responses will depend

16

the specific predisposition, the relative risk, the practicality of modifying
the enhanced risk for the predisposed worker (and possibly for all other workers
at the same time) by altering the environment, or altering work prattices, or
altering associated risk factors. The most important real-life test of how

we might respond to a specific risk factor is the major contribution of cigarette
smoking to so many occupational health problems. When we consider so-called
"hypersusceptible" workers, the most compelling group of workers--and a large
proportion--is the smoker. We should emphasize also the apparent adverse
effects to co-workers who are not smokers both from the smoke and its contents
and from the increased accident rates. The most desirable means of addressing
the smoking problem, in my view, is through a vigorous health education campaign
in which unions and management voluntarily participate and in which individual
workers can be encouraged to take steps to enhance their own health.

Your question about screening serving as a substitute for improvements in workplace quality is central to the labor/management tension over responsibility for workplace safety. As Elmes has stressed, however, substantial improvements in work conditions in the coal, iron and steel, asbestos and slate industries have had far less overall results than would be feasible if reduction of smoking had accompanied these environmental changes. It is my impression that the Johns Mansville Company policy to assist workers in certain jobs to stop smoking or switch jobs has been supported in arbitration rulings primarily because the company made very substantial investments in reducing exposures and did actually control most of the asbestos dust exposures. The genetic predispositions--if and when screening tests are developed--will serve to identify workers for whom the usual surveillance program may be an evaluation of the adequacy of environmental controls to protect highly susceptible workers. If adequate, such information should be reassuring also to less susceptible workers. As in the Johns Mansville smoking policy, however, there will come a point at which it may be infeasible technically or too costly by some criteria to offer environmental protection. Personal preventive health measures or job-switch would be required

17

Your final questions, still hypothetical, about whether workers should be permitted to assume risks, either health risks or liability risks, move into legal ground for which I have no background. I believe it would be difficult to prove that the worker seeking compensation had fully comprehended the "choice" made years earlier. In states with workmen's compensation programs that allocate responsibility there may be precedent for dividing the risks, but I doubt that we can find much precedent for informed consent for the kind of risk-taking you raise. In general, our society seems to tolerate an extraordinary degree of risk-taking by individuals, even when innocent others may be injured or be left parentless or when the costs of the consequences must be borne by the society. For so-called "involuntary risks", our social position has been that such risks should be eliminated entirely or reduced to as low levels as feasible. I recommend that some careful legal analysis and some detailed analysis of specific instructive potential cases be carried out in order to provide a sound response to your question.

There has been relatively little progress and almost no activity

in industrial screening for genetic traits during the past 5 or 10 years. Perhaps your hearing today, Mr. Chairman, will stimulate the development of the kinds of information required for intelligently advancing this work.

Mr. GORE. Well, I appreciate your concluding comment. It's our hope that we will play a constructive role in trying to focus attention to these troublesome issues before they reach a stage where they will be almost impossible to solve.

Do you agree with the view of the future that I have expressed regarding the role of HLA antigen screening and other forms of genetic screening?

Dr. OMENN. In general I do, but I think it's going to be much more slowly coming than you expect. The HLA story burst upon us 8 or 9 years ago. From a scientific point of view we thought we finally had got a handle on these arthritic conditions, but it turns out, despite hundreds and hundreds of papers redocumenting the same association, we really don't know any more now than we did then about what it is that is the underlying mechanism for the ankylosing spondylitis or certain other conditions.

The particular tests, some of which I summarized here, have been spoken about for possible use in screening for 15 years. Every 5 years or so there's a summary of the field for the Joint Conference on Occupational Health. I happen to be giving that summary tomorrow in Nashville, but the fact is that from reading what was stated 10 years ago and 5 years ago and tomorrow, there's been no progression. In fact, what we are saying now may have some chance of finally moving the field forward. We ought to move away from the focus on screening per se and get some sound information upon which to make judgments that will be valid scientifically and draw the interest and cooperation of workers and management for selection of appropriate tests.

I think it will be very important, when organized screening is initiated to make good choices, to screen for risks important to the workers, not just for marginal hazard, and to insist on tests which are highly reliable and do not have too many false positives or false negatives.

Mr. GORE. In light of the earlier discussion submitted by the representatives of the railroad industry, it seems clear that the decision to use or not to use marginal markers will be substantially affected by the legal framework within which these decisions occur and while you say 300 percent is an appropriate cutoff point, the closest I could get with the earlier panel was somewhere around 3 percent as an appropriate cutoff point. Although no specific figure was given, that was an impression I was left with. That's quite a range of difference and where statutes like the FELA are concerned you're probably going to see tests that you would consider marginal.

Dr. OMENN. Well, I urge you and your staff especially to review the appendix material in the testimony of Dr. Meyers. I did not have it in advance and I don't have a copy, but I was glancing at it while I was listening to his testimony. If I heard and read him right, I understand his company's study of the past year to claim that the increased prevalence of back disability in what they describe as a high risk group compared to the low risk group was not 3, 5, or 10 percent; it was 22-fold higher.

Mr. GORE. I understand that and I reviewed that appendix.