« AnteriorContinuar »
Mr. GORE. Thank you very much. I appreciate that. It's your conclusion that preemployment back X-rays should not be used; is that a fair summary?
Dr. ROCKEY. I think that the practice is widely abused.
Mr. GORE. And you worry about the discriminatory effect of the program as it's currently applied to prospective employees who are turned down when the predictive value is as low as it is?
Dr. ROCKEY. Yes, that's my major concern. I think that if the program were used after the decision had been made to hire the employee-an employee was already in the fold, so to speak-and then the X-ray was used as part of a placement procedure or part of a procedure to get the person into the best job, then it might be useful. But I think that the railroad industry has barred employment of these individuals based upon a medical test. However, I understand why they have done that, the FELA is especially repressive.
Mr. GORE. What's your opinion of the 2-percent or 5-percent hypothetical?
Dr. ROCKEY. I think that you have to look at it in the context of the social values that are being balanced. It's like the example that I used with the coronary artery disease. I don't think we would have wanted the pilot of the Space Shuttle to have any risk of coronary artery disease, but I think that if you're employing someone in your office building it's not your responsibility to determine whether or not they have coronary disease and discriminate in employment. So it's shades of gray.
Mr. GORE. Do you agree with my conclusion that the genetic screening potential is going to make these issues much larger than they now appear?
Dr. ROCKEY. Yes, I do. As a matter of fact, I think that we know enough now about some of the risks associated with some of the HLA antigens that I'm surprised some of them haven't been picked up by industry.
If I could just show a few more slides related to the HLA antigen. If we make the following assumptions about the histocompatibility antigen which used to be called B-27 but is now called HLA27. (Slide.]
ASSUMPTIONS The histocompatability antigen B-27 occurs in about 7 percent of the healthy white population.
About 95 percent of patients with ankylosing spondylitis are B-27 positive.
It occurs in about 27 percent of the healthy white population. It's variable in different populations. And about 95 percent of the patients with ankylosing spondylitis are positive for HLA-27 and the presence of HLA 27 predicts a risk of ankylosing spondylitis of 20 to 25 percent. Those assumptions are based upon several different studies. We would generate a similar table to what we have looked at before in the next slide. [Slides.]
The histocompatibility antigen B-27 occurs in about 7 percent of the healthy white population.
About 95 percent of patients with ankylosing spondylitis are B-27 positive.
75 to 100 percent of those refused employment would never have the disease.
INDUSTRY "COST-BENEFIT" ANALYSIS 1000 test at $50 per test-$50,000-“cost” to industry. 14 cases at $10,000 compensation per case-$140,000 “benefit” to industry.
If the disease prevalence is about 15 out of 1,000, HLA-27 screening would correctly identify 14 of those 15 individuals. That's the sensitivity of the test.
Mr. GORE. And this is a back problem of a kind that would be-
Dr. ROCKEY. Ankylosing spondylitis is a medical condition where the back fuses. It's quite different than what was being referred to. It's not the garden variety back problem that affects 60 percent. It affects only about 1.5 percent of the population. But again, that 1.5 percent really develops severe back disability.
Now if you look at what would happen if this test was used for preemployment screening. Because of the false positives, if we used this to exclude people from the workplace, we would exclude 70 people in order to get the 14 that would be at high risk and we would have misclassified 56. We would have told those 56 individuals that they were at risk for a disease that they will never get.
Mr. GORE. Well, I draw two conclusions from your example. First of all, it supports my suspicion that the potential for HLA antigen screening is immense because it is more predictive. Even in the specific example of railroad employment which we chose not for these reasons, but even here it's more predictive than the tests that are now underway. Would you agree with that?
Dr. ROCKEY. I would agree with that. If it were possible to genetically classify individuals and find out which ones were at risk for the early degeneration of the disks, that obviously would be the predictor that people would be interested in.
Mr. GORE. And my conclusion again is simply that that confirms my suspicion or tends to confirm my suspicion of the fantastic potential of this new technology, but my second conclusion is that in the confines of this specific example we still see present the potential for discrimination against healthy people who may not have the condition at all and discrimination against people who are classified inaccurately. So it's quite troubling.
I'm not sure how we're going to solve the problem, but it's good to get a start on it before it gets out of hand.
I'd like to thank you for your testimony here today. We are going to move on to our last witness, but I appreciate very much your taking the time to come and share your views with us and your slides particularly. Did you have anything to add before we move on?
Dr. ROCKEY. No. I think that the one point I wish to emphasize is that our best tests have sensitivities and specificities in the range of 95 to 99 percent and when the disease prevalence is low, even excellent tests will always result in a misclassification of individuals.
Mr. GORE. OK. Well, thank you very much. We certainly appreciate it.
Our final witness today in the first day of this 2-day hearing is Dr. Gilbert Omenn from the Brookings Institution and the University of Washington, Seattle. Dr. Omenn, we are pleased to welcome you back to the subcommittee. We appreciate your coming again and, without objection, your testimony will be put into the record in full and we invite you to proceed as you wish.
STATEMENT OF GILBERT OMENN, BROOKINGS INSTITUTION
Dr. OMENN. Thank you, Congressman Gore. In fact, it's a hectic morning. Two floors down from here there's a clean air hearing in which I'm also speaking.
I will present the highlights of the written testimony. I think this hearing has generated quite a fascinating discussion of the state of the art of diagnostic screening tests and also the need to be rigorous about assessing these kinds of medical technologies in any applications to the workplace.
There are three classes of reasons for screening in the workplace. They are to protect the worker, to support the regulatory process that's required under the Occupational Safety and Health Act, and to reduce the employer's liability. All of those are valid and necessary.
Despite the reasons given above and despite continued costly screening procedures, preplacement evaluation has fallen into con
siderable disrepute. Its benefit to either the employee or the employer has been challenged. In high-turnover, low-hazard industries, there may be a poor benefit-cost ratio. Some industries have approached this problem by reducing costs through use of paramedical personnel, especially nurses, to do the evaluations, which is generally quite satisfactory. Another reason for the disrepute is the often poor quality or lack of specific relevance of the evaluation and tests to the particular hazards in a given workplace. A third reason is the disparity in interpretation of findings by physicians or other health professionals.
Finally, there is the recognition that highly motivated individuals can often overcome seemingly serious physical handicaps. Laws to assist the handicapped have been beneficial.
I will direct the bulk of my comments to the specific area of genetic differences. There can be no doubt that people differ remarkably in many body functions and in the underlying genes and gene functions. Studies have been done for many years of individual variation in blood groups, plasma proteins, red blood cell enzymes, histocompatibility antigens, and now DNA sequences. At all levels of investigation, very significant variation has been demonstrated.
Advances in genetics have been truly remarkable, and there's considerable interest in the application to medicine.
There have been several kinds of genetic screening carried out. We should be explicit in distinguishing screening tests of patients, such as persons who already have back problems, from genetic screening of unsuspecting individuals in the population. The problem of informed consent is severe in the general screening situation. We have learned a lot, somewhat to our chagrin, from population screening for phenylketsonuria (PKU) and sickle cell disease and sickle cell trait.
One word about PKU has to be mentioned. I can assure you that seeing patients who are identified by screening as having metabolic defects of PKU in the early days of life and put on a special diet and grow up to be normal instead of being institutionalized for life with the mentally retarded, generates great pride in the technology that enables us to do that.
However, not too many people are aware that when the PKU screening was introduced-and the screening test here has 98 percent sensitivity and 99.9 percent specificity—the result was that the incidence of the disease apparently doubled. Twice as many newborn people were found with high levels in the blood and were stamped as PKU as were expected previously from counting bodies. That seemed to indicate that we were missing them before. Wrong. We were misdiagnosing an equal number to the number properly diagnosed as PKU. Now we understand that the levels may be high in the blood and in the urine due to other reasons which are generally harmless. All of these are interesting scientifically, but in practice a lot of children were put on the PKU diet until we learned that we have to retest them after 3 months. Tragically, there were a few among those who had a variant of the enzyme which requires higher levels of phenylalanine; those kids do badly on a low phenylalanine diet. So a few kids suffered mental and physical retardation of growth as a result of the misdiagnosis. Thus, even in one of our very best examples there were problems.
In the case of sickle cell screening there's a much worse situation. There was confusion between sickle cell disease, which is a serious anemia, and sickle cell trait. Eight percent of the American black population has sickle cell trait. Unfortunately, some people confused the disease and the trait and many case reports appeared from good physicians identifying the joint occurrence of sickle cell trait and kidney infections, problems in pregnancy, or any of a host of complications. It's widely stated that people with sickle cell trait are predisposed to have these various conditions.
The only way that can be assessed is to do a population-based study, not case by case. You start with the recognition that 8 percent of the black population has sickle cell trait and a certain percentage of the black population has these other specific diseases or symptoms. You must find out if there is a joint occurrence which is greater than chance, including the bias of ascertainment that you're following the patients closely, which is very important.
After nearly all of those associations melted away, the only one that was left and has clear implications for our discussion today is the well-acknowledged fact that red blood cells containing sickle cell hemoglobin will distort and sickle if you put the cells in low oxygen. It was deduced and it is true that if you put people at very high altitude or airplanes lose their pressurization that persons with sickle cell trait would be at risk to have sickling of their cells, with pain and possibly more severe reactions to the pain or damage.
As a result, the Air Force had a policy that no black American who had sickle cell trait would be admitted to the Air Force Academy and no black with sickle cell trait would be permitted to take pilot training, whether they went to the Air Force Academy or not. That was challenged with the same kind of logic that was presented here by Dr. Rockey and with similar lack of success until, after 10 years, the Air Force last year was overruled by the Department of Defense and instructed to admit blacks with sickle cell trait to the Air Force Academy and to pilot training programs. The Air Force will carry out the kind of medical surveillance which we think would be appropriate, especially in this case where there's uncertainty and need for information.
This is an important example, one which was fought about in the Federal Government for quite a period of time with experts from both sides and from outside the Government. Unfortunately, these experts and policy officials were making decisions in the absence of evidence. For the future, it will be very important to have a systematic way of gaining information, asking the questions you're asking now and gaining the evidence so we can make decisions about what kind of screening truly protects workers and potential workers and which kind of screening is so marginal that it should not be utilized.
In addition to sickle cell trait, another test that has been described in the literature and utilized is the serum antitrypsin deficiency test for predisposition to chronic pulmonary disease. Pulmonary emphysema is characterized anatomically by destruction of the alveolar air spaces of the lung. The mechanism appears to be