ethics and medical genetics in the united states. a national survey

American Journal of Medical Genetics 29:815-827 (1988)

Ethics and Medical Genetics in the United States: A National Survey

Dorothy C. Wertz and John C. Fletcher

Health Services Section, School of Public Health, Boston University, Boston, Massachusetts (0. C. W.); University of Virginia Medical School, Charlottesville. Virginia (J. C. F.)

The approaches of 295 medical geneticists in the United States to 14 clinical problems and 3 screening situations that required a moral choice are summarized. These data are part of a survey of 682 geneticists in 19 nations. Of 490 U.S. geneticists asked to participate, 295 (60%) returned anonymous detailed questionnaires. There was strong (> 75%) consensus that preserving the mother’s confidentiality overrides disclosure of true paternity; that conflicting test results, new/controversial interpretations of results, and ambiguous/artifactual results should be disclosed; that artificial insemination by donor, adoption, taking chances, contraception, sterilization, and in-vitro fertilization with a donor egg should be presented as reproductive options to carriers of disorders not diagnosable prenatally; that prenatal diagnosis should be performed for patients who refuse abortion and for maternal anxiety in the absence of medical indications; that screening in the workplace should be voluntary. There was no consensus about disclosure of a diagnosis of Huntington disease or hemophilia A to relatives at risk, against the patient’s wishes, or about disclosure of parental translocations. Geneticists in the U S . differed from 18 other nations in presenting surrogate motherhood as an option (67%); willingness to perform prenatal diagnosis for sex selection or refer (62%); and disclosure of XY genotype in a female (62%). Men were more likely than women to say that they would give directive counseling. Women were more likely than men to say that they would perform prenatal diagnosis for maternal anxiety or for sex selection.

Key words: prenatal diagnosis, sex selection, genetic counseling, disclosure, confidentiality, genetic screening

Received for publication July 13, 1987; revision received September 16, 1987.

Address reprint requests to Dr. Dorothy C. Wertz, Associate Research Professor, Health Services Section, School of Public Health, Boston University, 80 East Concord Street, Boston, MA 021 18.

0 1988 Alan R. Liss, Inc.

816 Wertz and Fletcher

INTRODUCTION

New tests, techniques, and treatments have increased the complexity of ethical decision-making in medical genetics. In the future, medical geneticists’ views and approaches will clearly be of great importance as clinicians, policy makers, and patients struggle to resolve ethical problems. Individual geneticists in several nations have set forth their ethical views [Berg, 1983; Crawfurd, 1983; Czeizel, 1983; Pfeiffer et al, 1983; Schroeder-Kurth, 19841, but there has been no systematic study of the actual approaches of medical geneticists to ethical problems. Fletcher et al. [1985] proposed that medical geneticists would benefit from collective reflection on their preferred approaches to the most difficult moral choices. In response, we studied the degree of consensus and variation among medical geneticists, when they were presented with ethical dilemmas concerning complex genetic situations. Here, we summarize findings on the approaches of 295 medical geneticists in the Unites States to 14 clinical problems and 3 screening situations that required a moral choice. The data presented here are part of a cross-national survey of 682 geneticists in 19 nations.

METHODS

In developing the questionnaire, we composed 14 clinical cases involving ethical problems in clinical genetics. These were developed on the basis of fieldwork by J.C.F. at 25 genetic centers in 12 nations. Respondents were asked what they would do, from a fixed list of possible responses, and why, in their own words, they had chosen this particular course of action. They were instructed to cite ethical reasons rather than purely technical or legal reasons for their actions. They were also asked which cases had given them the most and the least ethical conflict. There were 3 questions on screening for genetic disorders. Next, the respondents were asked to place 10 issues in rank order, beginning with the one of most ethical concern. The questionnaire was pilot-tested twice, on 11 and 10 geneticists respectively, from the United States and Canada, and revised after each test.

The questionnaires took approximately 2 hr to complete. We asked respondents to set them aside for a day after completing the first half, so that they might approach the second half freshly. Questionnaires were answered anonymously.

Criteria for Selection

We selected geneticists who held an M.D., Ph.D., or equivalent degree, and were engaged in delivering or administering genetic services (testing, counseling, prenatal diagnosis, laboratory work). Although counseling in the United States is sometimes done by specially-trained persons who do not hold a doctorate, we decided to omit these persons to control for consistency of training across the entire international sample.

We tried to include all qualified medical geneticists in the survey. Lists were compiled from the 1982 membership directory of the American Board of Medical Genetics and the National Foundation-March of Dimes “International Directory of Genetic Services,” 1983.

Of the 490 United States geneticists asked to participate, 295 (60%) returned completed questionnaires by the close of the study in January, 1986. Eighty-eight percent

Ethics and Medical Genetics 817

of these answered all parts of the questionnaires, including stating in their own words why they had chosen particular courses of action. Seventy-four percent held M.D.s, 22% Ph.D.s, and 4% held other degrees. They had a median of 14 years in the practice of genetics; 86% were members of the American Society of Human Genetics, and 83% were board certified. Respondents spent a median of 45 hr a week in genetics. Sixty-seven percent were male, and 81% were married with a median of 1.5 children. Religious backgrounds were 46% Protestant, 28% Jewish, 14% Catholic, 6% other, and 6% none. As a whole, they were nonpracticing, attending a median of 3 religious observances a year, although 39% attended once a month or more. Fifty-one per cent characterized themselves as politically liberal, 18% as conservative, and 31% as both equally. Socio-demographical- ly, United States geneticists differed from respondents in most other nations on only one variable: attendance at religious services. In the United States, 39% attended more than once a month. The 2 other countries with more than 25% were France (63%) and India (35%). A comparison between 274 respondents and 208 nonrespondents listed in the 1986 combined “Membership Directory” of the Genetics Society of America, American Society of Human Genetics, and American Board of Medical Genetics showed no statistically significant differences between respondents and nonrespondents in type of degree, gender, geographical area, or subspecialty. Qualitative reponses, including the responders’ first 2 reasons for their actions and their anticipation of the consequences of their choices, were quantified, using a coding system developed by the authors. Responses were entered into a Statistical Package for the Social Sciences (SPSS-X) program.

In order to see whether geneticists’ responses were related to factors in their professional or personal backgrounds, we entered all socio-demographic data, including degree, age, gender, years of experience, patients per week, subspecialty, political inclination, religious background, and religiosity into a stepwise logistic regression for each of the 14 clinical cases, using their responses to that case as the dependent variable. This method orders each background variable in terms of its strength of association with the dependent variable, while controlling for other statistically significant (p < .05) variables. In addition, this analysis provides an estimate of the odds ratio, e.g., the odds that a geneticist with a particular background variable will choose a particular response to a clinical situation.

RESULTS

The 14 clinical cases covered 6 types of ethical problems frequently encountered in the practice of medical genetics. 1) Patient confidentiality versus duties to relatives at risk (Cases 1 and 2 plus one screening situation [2A]; Table I); 2) Full disclosure of sensitive information (Cases 3-5; Table TI); 3) Full disclosure of test results (Cases 6 8 ; Table 111); 4) Presenting reproductive options nondirectively (Case 9 for male and Case 10 for female carriers; Table IV); 5 ) Indications for prenatal diagnosis (Cases 1 1-13; Table V). Case 14 (not shown in Tables) involved counseling about fetuses with disorders commonly regarded as having a low burden, XO and XYY.

The cases described in Table I present a conflict between 2 acknowledged principles in medical ethics: the duty to respect patient confidentiality and the duty to warn third parties (relatives at risk for genetic disorders) of harm.


TABLE 1. Patient Confidentiality Versus Duties to Relatives At Risk*

Percent responding

Disclose Refer to Respect Disclose even if patient’s

Clinical situation confidentiality if asked not asked physician

I . Hemophilia A. A client with a child recently diagnosed with hemophilia A refuses to permit disclosure of the diagnosis and relevant genetic information to her relatives who may be at risk for con- ceiving children with hemophilia A. 36 25 29 10

2. Huntington disease. A client recently diagnosed as having Huntington disease (HD) refuses to permit disclosure of the diagnosis and relevant genetic information to siblings who may be at risk for Hun- tington disease. 39 29 24 8

2A. Presymptomatic test for HD. When a 99% accurate presymptomatic test for Huntington disease is developed that applies in all families, should results of the test be disclosed to any of the following against the patient’s wishes?

67 33 N/A Spouse 64 12 24 N/A Relatives at risk for HD 56 20 24 N/A

- Patient

*n = 295. N/A = not applicable

TABLE 11. Full Disclosure of Sensitive Information*

Clinical situation

Percent responding

Do not Disclose disclose

3. False paternity. You are evaluating a child with an autosomal recessive disorder for which carrier testing is possible and accurate. In the process of testing relatives for genetic counseling, you discover that the mother and half the siblings are carriers, whereas the husband is not. The husband believes that he is the child’s biological father. 6 94

4. Parental translocation. You identify a parent of a Down syndrome child as having a balanced translocation. 62 38

5. XY female. A woman undergoes diagnosis for infertility. Tests reveal that she is chromosomallv male (XY). 64 36

*n = 295.

In Cases 1 and 2, the patient (with Huntington disease [HD] or hemophilia A) has refused to permit disclosure of the diagnosis to relatives at risk for the same disorders. The potential benefits of disclosure are very different for the 2 disorders. HD causes mental degeneration, is untreatable (except for symptomatic relief and counseling), and is of late onset. Although presymptomatic diagnosis through DNA linkage analysis is now possible


TABLE III. Full Disclosure of Test Results

Percent responding

nondirective Partial No Full disclosure,

Clinical situation counseling disclosure' disclosure

6. Conflicting test results. Maternal serum alpha-fetoprotein has been elevated in your patient on two occasions, but level I1 ultrasound discloses no abnormality, despite careful examination of the fetal head, spine, abdomen, and kid- neys. The fetal karyotype is normal. Amniotic alpha-fete protein is elevated and acetylcholinesterase is borderline. These results raise the possibility of a small neural tube de- fect. What do you tell the patients?

7. New/controversial interpretations. Repeated maternal serum alpha-fetoprotein tests reveal a value that is below the norm. Although some studies have found low maternal serum alpha-fetoprotein values to be associated with Down syndrome, geneticists are not in agreement about how a low value should be interpreted. (Case description repre- sents state of scientific knowledge in June, 1985).

8. Ambiguous/artifactual results. Laboratory analysis of amniotic fluid cells suggests that the fetus may be a tri- somy 13 mosaic. There is disagreement among the medical geneticists responsible for the analysis as to whether or not these are artifacts of culture, in other words, false positives. Given the present state of knowledge, there is no way of resolving this disagreement scientifically within the legal time limit for termination of pregnancy, because the results of repeat tests will not be available until after 24 weeks ges- tational age. You were not responsible for the laboratory work in this case and have not taken one side or the other.

95

89

3

8

2

3

You are, however, the medical geneticist responsible for dealing directly with the prospective mother. 75 24 1

'Partial disclosure means disclosing possibility of an abnormality, without revealing that results conflict, interpretations are new/controversial, or that results may be artifactual and colleagues disagree.

in some families, it requires the cooperation of several family members, including one who is affected. For many individuals, HD still cannot be diagnosed prenatally or presymptom- atically. The clinical diagnosis may occur too late for disclosure to affect the reproductive decisions of the patient's siblings or children, and disclosure may only cause psychological distress. In contrast, hemophilia A is diagnosable prenatally and is treatable; disclosure could help relatives in reproductive planning. We expected that more would disclose a diagnosis of hemophilia A than of HD. This was not the case. Instead, 53% would tell the relatives of the HD patient, and 54% would tell the relatives of the hemophilia A patient. These percentages include 24% and 29% respectively, who would seek out and tell the relatives even if they did not ask for information. Thirty-nine percent would preserve the confidentiality of the HD patient and 8% would refer the matter to the patient's family physician for decision; 36% would respect the confidentiality of the hemophilia A patient


TABLE N. Pnsenting Reproductive Options* Percent rasponding

Discuss Discuss pros and cons, Discuss, only if

Clinical situation no advice give advice asked

9-10. Presenting options to male (9) and female (10) carriers of disorders not diagnosable prenatally. Evaluation of a child produces findings consistent with a diagnosis of tuberous sclerosis. Upon examining the parents, you find evidence that one carries the tuberous sclerosis gene, even though intelligence seems normal. After a discussion of the risk of having another child with tuberous sclerosis who might be severely affected, the couple asks you whether recurrence of the disorder can be prevented. The options are:

Artificial inseminationdonor 96 2 2 Adoption’ 95 2 3 Taking their chances” 88 10 2 Contraceptiona 85 8 7 Sterilization’ 84 4 12 IVF with donor egg 83 1 16 Surrogate mother 61 2 31b

*n = 295. “Responses were the same for male and female carriers. bIncludes 9% who would not discuss even if asked.

and 10% would refer to the family physician. In both cases those who would disclose were significantly more likely (p < .OOOl) to envision and discuss the consequences of their action in their comments on the questionnaires than were those who would preserve patient confidentiality.

In their reasoning about the hemophilia A case, 37% cited the relatives’ right to know, 34% the patient’s right to privacy, 22% the duty to warn third parties of harm, 23% preserving the doctor-patient relationship, 19% reproductive planning, and 8% the prevention of birth defects. Reasoning in the HD case followed a similar pattern, but fewer (2%) mentioned preventing birth defects and 6% said that disclosure might cause psychological harm to the relatives. The older the geneticist the more likely s/he would disclose the HD diagnosis; the odds of disclosure increased by 1.9 for each 10 yr of age. Respondents were more willing to respect patient confidentiality for presymptomatic tests for HD (one of the 5 screening situations) than they were for an actual diagnosis; 67% would not disclose results to patients against their wishes, 64% would not disclose to the spouse, and 56% would not disclose to relatives at risk. In their reasoning, 62% said that patients had a “right not to know” their test results if they SO chose. Many of those who would preserve patient confidentiality argued that in the event that accurate tests become available, the relatives could be tested themselves without needing access to anyone else’s results.

As regards disclosure of sensitive information (Table 11), there was consensus (94%) not to disclose false paternity to the husband (Case 3). Most (84%) thought that they had fulfilled their duties as geneticists by telling the mother alone, without the husband present. By so doing, they have informed her that she is a carrier and that the disorder will


TABU V. Indications for Prenatal Diagnosis*

Percent responding

Favor no regulations

against Refer routine

to someone performance who would by commercial

Clinical situation Perform perform Refuse lab

11. Parents would refuse abortion. A 42-year-old woman requests prenatal diagnosis for Down syndrome. She and her husband already have a Down syndrome child. She tells you that they are opposed to abortion and that she will carry the fetus to term even if it is diagnosed as having Down syndrome. They would like to have prenatal diagnosis, however, in order to give themselves time to prepare for the birth of another affected child.

12. Maternal anxiety. A 25-year-old woman with no family history of genetic disorders and no personal history of exposure to toxic substances requests prenatal diagnosis. There are no genetic or medical indications for its use. Nevertheless she ap- pears very anxious about the normalcy of the fetus, and persists in her demands for prenatal diagnosis even after being informed that in her case the po-

96

tential medical risks for the fetus, in terms of mis- carriage, may outweigh the likelihood of diagnosing an abnormality. 78

0

11

4

11

90

80

13. Sex selection in absence of X-linked disease. A couple requests prenatal diagnosis for purposes of selecting the sex of the child. They already have four girls and are desperate for a boy. They say that if the fetus is a girl, they will abort it and will keep trying until they conceive a boy. They also tell you that if you refuse to do prenatal diagnosis for sex selection, they will abort the fetus rather than run the risk of having another girl. 34 28 38 50

*n = 295.

not occur in offspring sired by her husband. Ten percent would lie (tell the couple that they are both genetically responsible or that there is a new mutation) in order to preserve the mother’s confidentiality. The major reasons given for protecting confidentiality were preserving the family unit (53%) and the mother’s right to decide what to do with the information (34%). Only 11% thought that she ought to tell her husband. The odds that women geneticists would preserve the mother’s confidentiality were 6.6 times those for men geneticists, and the odds that women geneticists would mention marital conflict in their comments on the questionnaries were 2.7 times those of men.

In Case 4, parental translocation, genetic testing has revealed which parent carries a balanced translocation that has caused Down syndrome in their child (Table 11).


Disclosure of this information might enable the couple and relatives at risk to use reproductive options that would prevent the birth of another Down syndrome child, but could also cause guilt in the carrier or threaten the marriage.

In all, 62% would disclose, unasked, which parent carried the translocation. This included 14% who would also attempt to locate and disclose to all relatives at risk of being carriers. Thirty-seven percent would tell the couple that the test results showed which one was a carrier, and let them decide whether they wished to know; this included 25% who would tell them before the test that carrier status would be disclosed. Only 1% would wait for them to ask, and none would conceal the results. In their reasoning, 42% said that the couple had a right not to know, and 16% said that they had a duty to know and to use the information. Women were 1.9 times more likely than men to cite the welfare of relatives at risk.

In Case 5 , a phenotypic woman presents for infertility and is found to have XY genotype (Table 11). Disclosure could severely damage her self-image, but could resolve doubts about fertility. As in the previous cases, there is a conflict between the physician's duty to tell the truth (which may also be phrased in terms of the patient's right to know) and the duty to do no harm. The patient has not asked for the information and is unaware of its existence. Sixty-four percent would disclose XY genotype, and 36% would give other reasons for infertility. Thirty-three percent believed that they could tell the truth in such a way as to minimize harm, by providing supportive counseling, 32% saw the truth as a source of harm, and 29% would tell the truth in order to avoid the harm that would result from the patient's learning the truth from someone less skilled in counseling. In both this and Case 4, those who would not disclose were significantly more likely (p < .0001) to envision and discuss the consequences of their actions than those who would disclose.

There was widespread agreement that conflicting, new/controversial, or ambiguous/artifactual test results should be disclosed without giving directive advice (Table 111). The only exception was disclosure of colleague disagreement in the case of ambiguous/ artifactual results. Although 99% would disclose the possibility of an abnormality, 24% would not tell the patient that their colleagues disagreed.

Table IV describes the presentation of 7 reproductive options, artificial insemination by a donor (AID), adoption, taking chances, contraception, sterilization, in-vitro fertilization (IVF) of a donated egg, and insemination of a surrogate mother with the husband's sperm, to carriers of a serious genetic disorder not diagnosable prenatally (tuberous sclerosis). The percents listed are those who would present each option without being asked, and who would discuss it at length without giving directive advice. Table IV documents wide acceptance of AID; 96% would present this as an option for male carriers. Counseling of male and female carriers was the same with regard to 4 options available to both: 95% would present adoption, 88% taking their chances, 85% contraception, and 84% sterilization. IVF with donor egg has thus far produced about a dozen live births around the world, but 83% would present this as an option. Sixty-seven percent (72% of men and 57% of women) would present surrogacy as an option, 22% would discuss it if asked, and only 9% would refuse to discuss it at all. Many respondents said they they regarded IVF with a donor egg as ultimately less complicated then surrogacy and less likely to cause harm.

Table V describes 3 cases in which patients either request prenatal diagnosis without medical indications or say that they will not abort an affected fetus. In Case 1 1, a couple aged 42 with a Down syndrome child requests prenatal diagnosis in order to prepare themselves for the possible birth of another Down syndrome, but say that they will not


terminate the pregnancy. Ninety-six percent of respondents would perform prenatal diagnosis for this couple. In their reasoning, 69% stated that performance of prenatal diagnosis should not depend on the use that patients intend to make of the information. Thirty-four percent stated that such patients may change their minds about termination and thereby justified performing prenatal diagnosis. Refusals were based on lack of resources.

In Case 23, a woman of 25 with no personal or family history of genetic disorder or toxic exposure requests prenatal diagnosis because she is extremely anxious about the health of the fetus. Seventy-eight percent would perform prenatal diagnosis for maternal anxiety in the absence of other indications and 11% would refer the patient to someone who would perform it. Of those in favor, 52% mentioned patient autonomy and 32% mentioned the removal of anxiety. Only 14% cited possible harm to the fetus from the procedure. The odds that women would perform prenatal diagnosis for an anxious woman were 2.8 times the odds for men; 51% of women and 33% of men cited patient autonomy in their reasoning.

Case 13 (Table V), use of prenatal diagnosis solely for selecting the sex of the child, in the absence of an X-linked disease, was the most controversial in the entire questionnaire. In the case description, parents of 4 healthy daughters desire a son and threaten to terminate the pregnancy unless prenatal diagnosis is performed for sex selection, rather than risk having a fifth girl. If prenatal diagnosis is performed, they will terminate only if the fetus if female. Thirty-four percent would perform prenatal diagnosis for sex selection, 28% would refer the couple to another medical geneticist or genetics unit offering the service, and 38% would refuse.

Of those who would perform prenatal diagnosis, 68% said that they would do so out of respect for parental autonomy; and 19% would do so to prevent the otherwise certain abortion of a normal fetus. Those who would actually perform prenatal diagnosis were more likely (p < .0001) to set down the consequences of their actions, whereas those who would refuse or refer did not give their rationale. Stated consequences related to the fetus or parents, and not to society. Only 3.4% mentioned the position of women in society, 3.3% mentioned maintaining a balanced sex ratio, and 4% mentioned setting a precedent that would harm the moral order. Women were twice as likely as men to say that they would perform prenatal diagnosis; women would do so on the basis of respect for patient autonomy.

In connection with the 3 prenatal diagnosis cases, we asked whether commercial laboratories should be regulated. Our question was phrased, “A large commercial laboratory plans to open soon in your area. This lab has announced that it will have associated board certified obstetricians on the premises and that it intends to perform prenatal diagnosis for anyone who desires it and who is willing to pay the fee. Should there be any regulations prohibiting this lab from performing prenatal diagnosis routinely for the following indications?”

In all, 90% favored no regulations prohibiting performance for parents who refuse abortion, 80% favored no regulations prohibiting performance for maternal anxiety in the absence of medical indications, and 50% favored no regulations about sex selection (Table V). Twelve percent said that ethics should not be established through regulations; only 2% cited the “Relman Rule” that “medicine should not be a business.”

In Case 14, fetuses with low burden disorders XO and XYY (not shown in table), 90% and 9596, respectively, would counsel nondirectively; 24% would also include a discussion of the emotional difficulties associated with terminating the pregnancy. Women


were 4.4 times more likely than men to say that they would counsel nondirectively about XYY fetuses, and 3.6 times more likely than men to counsel nondirectively about XO fetuses.

Most and Least Difficult Cases

The cases that respondents found most difficult to resolve were sex selection (28%), confidentiality of an HD patient (18%), and false paternity (13%). In all, 50% said that these cases were the most difficult because of conflicting responsibilities to different parties, and 43% described conflicts between ethical principles. The least difficult cases were prenatal diagnosis for parents who refuse abortion (26%), prenatal diagnosis for maternal anxiety ( 1 4%), and disclosure of parental translocation (1 3%).

Screening The questionnaire included 3 screening situations: screening in the workplace for

serum a- 1-antitrypsin (SAT) deficiency, carrier screening for cystic fibrosis (CF), if and when available on a population-wide basis, and presymptomatic testing for HD, discussed above. As yet, SAT testing has not proved to be predictive of the development of emphysema. Seventy-seven percent thought that screening in the workplace should be voluntary; 34% cited potential conflicts between workers and employers, and 29% mentioned stigmatization. All thought that the worker should be told the results of any such screening; 24% thought employers should have access without consent, and less than 12% thought that insurance companies should have such access.

There was no consensus about the best approach to a hypothetical test that would detect both CF carriers and homozygotes; 23% said the test should be given to newborns by law; 23% to newborns with parental consent, 13% to adolescents ages 13-17, by consent, 35% preferred persons over 18, by consent, and 6% mentioned other ages by law. Only 16% thought that screening should be restricted to Caucasians; most considered such a limitation racially discriminatory.

Total Consensus

There was 275% consensus on 9 (64%) of the 14 clinical cases regarding: 1) non-disclosure of false paternity; 2) disclosure of conflicting test results; 3) disclosure of new/controversial interpretations; 4) disclosure of ambiguous/artifactual results, including colleague disagreements; 5 and 6) presenting reproductive options to carriers of disorders not diagnosable prenatally (except surrogacy, 67%); 7) performance of prenatal diagnosis for parents who refuse abortion; 8) performance of prenatal diagnosis for maternal anxiety; and 9) nondirective counseling about fetuses with low-burden disorders. There was 275% consensus on one of 3 screening situations, screening in the workplace and access to results.

Future Priorities

We asked respondents to rank-order, from one to 10, a list of future issues that they thought should be of most concern to medical geneticists in the United States in the next 10 to 15 years. They ranked these issues as follows: 1) development of new treatments for genetic disorders, including treatment in utero, organ transplantation, and molecular genetic manipulation; 2) carrier screening for common genetic disorders; 3) increased demand for genetic services; 4) environmental damage to the unborn; 5) allocation of limited resources; 6) screening for genetic susceptibility to cancer and heart disease; 7)


research on the human embryo, zygote, and fetus; 8) genetic screening in the workplace; 9) long-range eugenic concerns; and 10) sex preselection for sex desired by parents.

How the U.S. Differs From Other Nations

A total of 682 geneticists (64% of those asked to participate) in 19 nations (Australia, Brazil, Canada, Denmark, Federal Republic of Germany, France, German Democratic Republic, Greece, Hungary, India, Israel, Italy, Japan, Norway, Sweden, Switzerland, Turkey, United Kingdom and United States) responded to the survey. Results for other nations have been reported elsewhere [Fletcher et al., 19871. There were 4 questions on which the United States stood virtually alone among nations: surrogate motherhood, sex selection, disclosure of XY genotype in a female, and regulations for commercial laboratories. The United States was the only nation where a majority of geneticists (67%) would present surrogate motherhood as an option in counseling. The United States was one of 3 nations where a majority (62%) would either perform prenatal diagnosis for sex selection or refer to someone who would. In the United States, those who would perform would do so out of respect for parental autonomy. In the other 2 nations, Hungary and India, reasons were different. Sixty percent of Hungarian geneticists would perform prenatal diagnosis to prevent the abortion of a normal fetus, and 52% of Indian geneticists would perform or refer for social reasons, such as limiting the population or preventing harm to an unwanted girl. The United States and Canada were the 2 nations where more than 62% would disclose XY genotype. Willingness to disclose in these nations was based on confidence that, with proper counseling methods, the truth could be told in a manner that would not destroy the woman’s self-image. This approach reflects increases in the numbers of geneticists who have undergone clinical training in counseling. Finally, the United States stood alone in regard to commercial laboratories’ performance of prenatal diagnosis for sex selection, with 50% favoring no regulations. Among the other nations, Canada came closest, with 34% favoring no regulations; elsewhere 282% said that regulations should prohibit performance.

DISCUSSION

We expected to find a consensus in approach, based on the following policies: 1) respect for patient and parental autonomy, including an obligation to respect patient requests with which the geneticist disagrees; 2) duty to reduce or prevent suffering from genetic disorders, conditional upon respect for parental autonomy (e.g., counseling should be non-directive); 3) full disclosure of test results, including colleague disagreements; and 4) a voluntary, not mandatory, approach to genetic screening, except for newborns when treatment is available. We found consensus with regard to many of these practices, but not all. In the United States, geneticists have gone to greater extremes than any other country in acceding to patient requests. Two of the 4 questions on which the United States differed from all other nations (surrogacy and sex selection) represent fulfillment of requests, even if the geneticist had moral reservations. Responses to a third question (commercial labs) indicated support for unregulated free enterprise and for allowing patients to have access to whatever services they can pay for, a situation that perpetuates inequality.

Sex selection is clearly the most controversial ethical issue at present in medical genetics, when viewed in terms of a growing trend to meet parents’ requests for this purpose. Responses from the United States contrast markedly with earlier findings that


only 1% in a sample of 448 [Sorenson, 19761 and 21% in a sample of 149 geneticists [Fraser and Pressor, 19771 were willing to perform prenatal diagnosis for sex selection only.

As chorionic villus samping (CVS) becomes available, medical geneticists may become even more willing to comply with this controversial request. In the United States, CVS is still a new technology; it is difficult to predict how many patients will seek to use it solely for sex selection [Steinbacher and Holmes, 19871. Even though the actual number of requests may be few, the larger ethical question of harm to the health care system from diversion of resources remains. Many geneticists appeared to regard resources as unlimited and would perform prenatal diagnosis on request without genetic indications, unaware that by so doing they might be depriving patients with clear genetic indications. Overall, geneticists seemed to have little awareness of the social consequences of sex selection. TO some, it appeared to be an extension of families’ self-evident right to determine the number, spacing, and quality of their children. Others mentioned the ease with which patients desiring sex selection could conceal their real reasons for requesting prenatal diagnosis, for example, by claiming that they had toxic exposure. Although geneticists could prevent the use of prenatal diagnosis for sex selection by withholding information on fetal sex, this contravenes the ethic of full disclosure and respect for patient autonomy. It is doubtful whether withholding this information would be tolerated by patients in the United States with its fee-for-service health system, [Hulten et al., 19871.

Geneticists’ rankings of future priorities suggest that they put low priority on issues that are of great concern to the public, namely sex selection and genetic screening in the workplace. Future actions on these issues could change the entire structure of society by altering family composition or limiting access to work. Although in the United States most families would prefer to have children of both sexes, the majority would prefer that a first-born or an only child be a boy [Gilroy and Steinbacher, 1983; Pharis and Manosevitz, 19841. Sex selection technology thus has a potential for changing the birth orders and social relationships of males and females. Mapping the human genome, together with improved techniques of DNA linkage analysis, could bring eugenic concerns, also given low priority by geneticists, quickly to the forefront.

Societal problems tend to sneak up on medicine and catch physicians unaware. The responses to ethical problems by medical geneticists reflect a general confidence about resolving moral conflicts at the individual-familial level and a lack of will to make a contribution at the societal-political level, especially in terms of questions that involve society’s interests. One explanation for this situation is that many geneticists avoid any appearance of eugenic considerations and shy away from societal issues, partly because of the stigma caused by the notion of eugenics. The emergence of increasingly powerful scientific and diagnostic tools, such as DNA probes, will not allow clinical geneticists to avoid the social implications of human genetics. In our view, medical geneticists need to consolidate their best insights of the past, so that they can be prepared to contribute to the development of improved and informed ethical guidance for the complex problems of the future.

ACKNOWLEDGMENTS

This study was supported by the Medical Trust, one of the Pew Foundation Memorial Trusts, administered by the Glenmede Trust Company, Philadelphia, PA, by the Muriel and Maurice Miller Foundation, and by the Norwegian Marshall Fund.


REFERENCES

Berg K (1983): Ethical problems arising from research progress in medical genetics. In Berg K, Tranay KE

Crawfurd M d’A (1983): Ethical and legal aspects of early prenatal diagnosis. Br Med Bull 39:31&314. Czeizel A (1987): “The Right to be Born Healthy: Ethical Problems of Human Genetics in Hungary.” New

Fletcher JC, Berg K, Tranay KE (1985): Ethical aspects of medical genetics. Clin Genet 27:199-205. Fletcher JC, Wertz DC, Sorenson JR, Berg K (1987): Ethics and Human Genetics: A Cross-Cultural Study in

17 Nations. In Vogel F, Sperling K (eds): “Human Genetics: Proceedings of the 7th International

(eds): “Research Ethics.” New York Alan R. Liss, pp 261-175.

York Alan R. Liss.

Congress of Human Genetics, Berlin (West), September 22-26, 1986.” Heidelberg: Springer-Verlag. pp 657612.

Fraser FC, Pressor C (1977): Attitudes of counselors in relation to prenatal sex determination for choice of sex. In

Gilroy F, Steinbacher R (1983): Preselection of child’s sex: Technological utilization and feminism. Psycho1 Rep

Hulten M, Needham P, Watt JL, Griffiths M (1987): Preventing feticide. Nature 325:190. Pfeiffer RA, FrCzal J, Giraud F, Anders G, Robert JM (1 982): Le g6nLticien confront6 aux probl&mes d’6thique

midicale. IXbme Journks Europknnes de Conseil Gtnktique, Erlangen, Sept. 1982. J Genet Hum 30 suppl 5:447466.

Pharis ME, Manosevitz M (1984): Sexual stereotyping of infants: Implications for social work practice. Soc Work Res Abstr 207-12.

Schroeder-Kurth TM (1 984): Ethische Probleme bei Genetischer Beratung in der Schwangerschaft. Monatsschr Kinderheilkd 1307 1-74.

Sorenson JR (1976): From social movement to clinical medicine: The role of law and the medical profession in regulating applied human genetics. In Milunsky A, Annas GJ (eds): “Genetics and the Law.” New York Plenum Press, pp 467-485.

Steinbacher R, Holmes HB (1987): Survival and sisterhood. In Corea G, Holmes HB, Steinbacher R (eds): “Man-Made Women: How New Reproductive Technologies Affect Women.” Bloomington, IN: Indiana University Press, pp 52-63.

Wertz DC, Fletcher JC (1988): “Ethics and Human Genetics: A Cross-Cultural Perspective,” Heidelberg: Springer-Verlag .

Williamson NE (1976): ‘Sons or Daughters: A Cross-Cultural Survey of Parental Preferences.” Beverly Hills, CA: Sage Publications.

Lubs HA, de la Cruz F (eds): “Genetic Counseling.” New York Raven Press, pp 109-120.

53:671476.

Edited by Philip Reilly

ethics and medical genetics in the united states. a national survey

Documents