Lives In The Balance: Assessing The Risks Of Waiting For Perfectly Accurate Tests

Discovery of the cystic fibrosis gene in late 1989 set off a flurry of excitement in the scientific community over the possibility of mass screening for the most common lethal recessive gene among persons of European descent. About one in 26 American whites carries the gene, and about one in 2,500 newborns has CF. Carrier testing would identify couples in which both partners carry the gene and would enable them to decide whether to take the one-in-four risk of having a child with CF or whether to pursue alternatives, such as adoption, donor insemination by a noncarrier, or prenatal diagnosis for CF, which is now possible. Yet, more than a year after identification of the gene, CF testing remains mired in controversy. As some geneticists foresaw years earlier, new tests are frequently accompanied by technical, economic, or social problems. The new CF tests have all three. They identify the F508 mutation, which accounts for about 76 percent of CF carriers among people of European descent, but the remaining 24 percent of carriers remain undiscovered. By testing for an additional three mutations, it is possible to discover about 84 percent of carriers in this population. For Ashkenazic Jews and African Americans, among whom the delta F508 mutation is a less common cause of CF, the tests identify a far smaller percentage of those who are actually CF carriers. The tests are not cheap: the usual cost to the patient for testing for four mutations is about $250, which is rarely reimbursed by insurance because the CF tests are still considered experimental. Further, no one is certain what the general population knows about CF or thinks about carriers. How many people confuse cystic fibrosis with multiple sclerosis? Or with mental retardation? How many think that children with CF still die before the age of 10? How many would avoid a known CF carrier in the office or at a party? The potential social problems occasioned by introducing tests on a massive scale for a little-known disorder could be horrific. The major controversy, however, centers on risk. Tempers have flared over the level of sensitivity that tests should meet before obstetricians offer them routinely as part of preconceptual or prenatal care. The problem is that at 75 percent sensitivity, tests will identify about 57 percent of the couples at risk for having children with CF. Even at 84 percent sensitivity, tests could identify perhaps 71 percent of couples. In either case, there will be couples left with uncertainties. Couples in which one partner is positive for one of the four mutations and one is negative still face a risk of one in about 650 of having a child with CF, whereas at 95 percent sensitivity, this risk is reduced to one in 1,964. Some geneticists worry that offering the test as it now stands may do more harm than good. For every 100,000 couples in the white population tested, 111 would be identified as carrier-carrier couples at a one-in-four risk, but about 2,228 couples would be given a risk of one in 650, a risk that cannot be resolved by further testing. Some argue that the harm from the uncertainty and the anxiety engendered by this unresolvable risk outweigh the potential benefits to the relatively few carrier-carrier couples identified. Weighing of risks and benefits is a serious ethical issue yet to be settled. Even the President's Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research has not provided guidance on actual numeric levels of sensitivity. A further problem is that people whose tests are negative are unlikely to return for repeat tests in the future, when tests will be able to identify more mutations. In view of these problems, the American Society of Human Genetics and some regional genetics groups have recommended that tests be offered to relatives of people with CF (who have a higher-than-average risk of being carriers), but not to the general population. Are these standards realistic? There are few absolute certainties in medicine. Many widely used tests, such as mammograms, identify perhaps two-thirds to three-quarters of the diseases they were intended to detect. Few would argue for discontinuing a diagnostic test for cancer on these grounds, and few doctors express concern that some tests leave patients with a false sense of security. Genetic screening, however, is not the same as testing for disease. Many people expect greater perfection from tests and treatments related to reproduction than they do from other areas of medicine. People want contraceptives to be perfect, with no side effects whatsoever; they desire "perfect" babies and become upset if there are no 100 percent guarantees. Medical geneticists are particularly sensitive to risk. For many years, genetic counseling was described as "restructuring uncertainty" because a major part of a geneticist's task was to present risks--not certainties--to prospective parents. A study of more than 1,300 genetic counseling sessions between 1977 and 1979 found that geneticists considered risks of 3 percent (the general population's risk of birth defects) or under as very low, risks of 4 to 6 percent as low, risks of 7 to 20 percent as moderate, and risks of greater than 20 percent as high. Patients, on the other hand, tended to see the same numeric risks as lower. Patients regarded risks below 15 percent as low, and risks of 15 to 50 percent as moderate. Estimations may have changed in the intervening years. An increased desire for perfect children may lead some couples to pay greater attention to small numeric risks. Nevertheless, geneticists who urge caution may be overestimating public reaction to the uncertainty inherent in currently available tests. Research on risks in general suggests that in the absence of 100 percent certainty, most people tend to overestimate small risks and to underestimate large ones. Numeric risk alone does not tell the whole story. A large part of risk interpretation depends on what is risked. Most people have a greater fear of catastrophic events with small numeric risks, such as nuclear power plant explosions, than of the ordinary events with far greater numeric risks, such as the consequences of smoking or of not wearing seat belts. Most people are probably oblivious to the general "population risk" of 2.5 to 3 percent for having a child with a birth defect. For individual families, some birth defects, especially those associated with severe mental disabilities, may fall into the category of catastrophic events, for which even a minuscule numeric risk is too high. CF, however, does not affect intelligence; the median lifespan of a CF sufferer is now about 26 and increasing as treatments improve. Many people with CF finish college, hold full-time jobs, engage in sports, marry, and become parents. If told these facts, will prospective parents consider a risk of one in 650 as catastrophic? Probably not. Many will be more concerned about the possibility of their regional nuclear power plant exploding, about toxic waste dumps, global warming, or loss of the ozone layer. Those who lie awake worrying over a one in 650 chance that a future child may have CF may be expressing wider emotional concerns and uncertainties about parenthood in general. Nevertheless, there is a very real minority for whom any level of uncertainty is too high. In a recent study of 222 parents of children with CF, 31 or 14 percent said a risk of one in 2,000 for CF was high or very high, and 52 (23 percent) said a risk of one in 100 was high or very high. (The latter figure dropped to 13 percent when the risk was phrased as 1 percent rather than one in 100.) Risk interpretation was not related to the CF child's health status, to the parents' future expectations for the child, or to demographic variables. If anything, it appeared to be a matter of individual psychology. Obstetricians claim that many complications of pregnancy carry risks of about one in 400, that a pregnant woman may face several such risks, and that some risks cannot be resolved by further testing. Therefore, they see nothing unusual in a genetic test that leaves some couples with a risk of one in 396. Instead, they see considerable benefit in enabling carrier-carrier couples to have prenatal diagnosis. Some obstetricians are already circulating letters among their colleagues urging that CF carrier testing be part of routine prenatal care. A counter-question to the geneticists' studied caution arises: What will actually happen if we wait until tests reach a desired level of sensitivity before offering them to pregnant women or to the population at large? In fact, market forces will take over. Objections by professional societies will not prevent obstetricians from routinely offering CF tests to their pregnant patients, in effect making tests a standard of care. Historically, other tests and procedures in obstetrics have become standards of care in spite of the recommendations of obstetrical societies. Ultrasound and fetal heart monitoring are two examples. The American College of Obstetricians and Gynecologists (ACOG) has never recommended routine use of either procedure, yet both became standard in pregnancy and birth without large-scale clinical trials. Fetal monitoring remains routine in spite of a 1988 ACOG recommendation that its use be limited to specified, high-risk circumstances. The desire for perfection--and the fear of lawsuits--are simply too great to reverse the trend. Malpractice suits usually turn on typical local practice, not the recommendations of national societies. Therefore, if several obstetricians in an area start to offer CF tests, others are likely to follow. Obstetricians have particular reason to fear malpractice suits; if birth defects are involved, the doctor can be held accountable for the child's lifetime care. ACOG surveys have suggested that "fear of lawsuit" is the major rationale for many obstetrical procedures. Under the current market system of medical care, however, not all pregnant women will be offered tests. Payment is a major issue. Tests currently cost patients about $250. As long as they are designated "experimental," most insurers and Medicaid will not pay for them. Testing will follow our two-class system of medical care. Patients with the education to request testing, the money to pay for it, and the savvy to sue will be tested; clinic patients will not. This contravenes the principle of justice. It could further increase social inequality by selectively enabling families with education and money to avoid the considerable financial and emotional burden of CF. If market forces do take over, there is no guarantee of uniform standards for physician and patient education about CF or CF testing. What doctors and patients do with the information from tests will depend largely on what they are told about CF. If doctor or patient thinks that most children with CF die before the age of 10 (which was the case two decades ago), they will view risks differently from someone who thinks life expectancy is 40 or 50 (a somewhat optimistic view shared by many parents of CF children) or someone who knows the median is about 26. Knowledgeable professionals and ethicists have agreed that testing must be accompanied by an impartial education program that has been thoroughly evaluated. If testing follows the piecemeal route through private practices on the way to becoming a standard of care, however, such a uniform program is unlikely. Scientists, doctors, and prospective parents face a dilemma. To proceed with routine testing now appears undesirable on account of the tests' limited usefulness. On the other hand, to wait perhaps for several years, until tests reach some desired but as yet unagreed-upon level of sensitivity, means that market forces will set the pattern for a two-class system of care. Pilot studies could help to resolve the dilemma. We need to find out what people know about CF, what they think about carriers, how they perceive and deal with risks, and how they can best be educated. To date, however, government agencies have refused to fund pilot studies. Soon the matter may be academic. To wait is in effect to opt for a takeover by market forces. Dorothy C. Wertz is a sociologist and research professor of health services at the School of Public Health at Boston University.

Lives In The Balance: Assessing The Risks Of Waiting For Perfectly Accurate Tests

Interested in reading more?

Become a Member of