Brief Overview of Constitutional Issues Raised by the Exclusion of Women from Research Trials
R. Alta Charo
The exclusion of fertile women from research trials, based upon both FDA regulation and the pattern of practice among private pharmaceutical companies and academic centers, is prima facie disparate treatment of two classes of persons: fertile females and all others. Whether the disparate treatment of these two classes raises constitutional issues depends upon a number of factors:
Does the disparate treatment result in any disparate benefits to the two classes?
If so, are these benefits classified as specially protected "rights"?
Whether or not the benefits are classified as "rights," is there a justification for the disparate treatment, in light of the disparate benefits?
If so, does the disparity in benefit operate to disadvantage a protected class of persons? If it does, the level of justification for the disparate treatment must be somewhat higher.
This author concludes that the exclusion of fertile women does have the effect of denying women as a class an equal opportunity to benefit from government-funded research. The justifications for the exclusion, which rest primarily upon calls to protect potential fetuses from harm or to minimize the costs of research, are only dubiously sufficient to sustain the exclusion. Exclusion of women from nongovernmentally funded research, however, is probably constitutional.
DISPARATE TREATMENT, DISPARATE BENEFITS, AND EQUAL OPPORTUNITY
1. Does disparate treatment result in any disparate benefits to the two classes?
The goal of excluding fertile women from research trials is often achieved by excluding all women or all premenopausal women from the trial (Nadel, M.V., "NIH: Problems in Implementing Policy on Women in Study Populations," GAO/Y-HRD-90-38). Even where sterile women are included, many are sterile owing to hysterectomy, thus rendering their hormonal patterns unrepresentative of the hormonal patterns in fertile women (Cotton, 1990). The degree to which the operation of various drugs and therapeutic protocols differs in women and men, or fertile women and all others, is unclear. Inadequate data exists, whether at in vitro, animal, or human levels, to accurately predict which interventions are likely to be significantly affected by the different average body weight, body composition, hormonal patterns, and collateral diseases or drugs used by men and women (Cotton, 1990).
Nonetheless, tantalizing data do suggest that these differences exist more often than had been suspected. This paper, which focuses on constitutional issues, will not review those data. (But see GAO, "Women's Health: FDA Needs to Ensure More Study of Gender Difference in Prescription Drug Testing," GAO/HRD-93-17, October 1992; SAWHR, 1991). Instead, it is written on the assumption that further research would reveal that many, but not all, drugs and therapeutic interventions have different degrees of risk and efficacy in male and female populations.
2. Are the benefits of being a research subject specially protected "rights"?
Robert Levine has amply demonstrated that there has been a paradigm shift in the way in which we view enrollment in research trials (Levine, 1988). While research on human subjects was first viewed as a necessary aspect of public health, and then viewed as a transgression of individual rights tantamount to torture, it has lately come to be viewed as an avenue of access to better medical care for oneself and one's cohort, 1991.
Thus, being a research subject is no longer viewed as an unqualified sacrifice. Rather, it is a potentially risky opportunity. The degree of risk depends upon the nature of the therapy being tested and the fragility of the subject. The degree of opportunity depends upon the extent to which the subject suffers from a disorder destined to be treated by the therapy, the availability of effective therapeutic interventions already on the market, the urgency of the subject's disorder, and the access the subject has to paid, therapeutic care.
Looked at this way, fertile women who are denied access to clinical trials
may or may companies or not be denied a benefit. The question of benefit is idiosyncratic. To the extent, however, that the exclusion of fertile women from research trials does bar access for some to potentially lifesaving treatments (IOM, "Expanding Access to Investigational Therapies for HIV Infection and AIDS," 1991), it is undoubtedly the denial of a benefit. Further, if the exclusion of fertile women results in all women, as a class, being denied adequate protection from the marketing of drugs and interventions dangerous to their health, then this exclusion denies all women an important benefit of research. In other words, by failing to adequately test drugs and interventions on women during the premarketing phase of development, the female consumers of those products in the postmarketing phase are transformed into unwitting subjects of a poorly controlled and poorly monitored large-scale, long-term research trial.
Although there is no "right" to health care or to the benefits of health research under the U.S. Constitution, there are specific rights to life and liberty that are implicated when women are denied potentially lifesaving interventions or are subjected to potentially life-threatening risks owing to the lack of appropriate research. Further, the Fifth and Fourteenth Amendments' guarantees of "equal protection" under the law, while not transforming "benefits'' into "rights," do treat the distribution of mere "benefits'' as a matter of constitutional significance.
3. Whether or not the benefits are classified as "rights," is there a justification for the disparate treatment, in light of the disparate benefits?
The usual justifications for excluding fertile women from research trials include: male-only data are more homogeneous and therefore more useful; inclusion of women in trials will be unduly costly; government regulation requires the exclusion of women; and the threat to potential fetuses creates a legal and moral imperative to exclude all potentially pregnant women (Merton, V., "The Exclusion of Women from Research Trials," presentation at the Texas Journal of Women and the Law Symposium: New Perspectives on Women, Health and the Law, 5–6 March 1993). Each of these justifications, however, can be shown to be seriously flawed.
Homogeneous data, for example, are only more useful when they adequately serve the purposes for which they are collected. If that purpose is to provide improved therapeutic interventions for men, then current research practices are entirely suitable. If, however, the purpose is to provide improved therapeutic interventions for all persons, then homogeneous data from male-only subject pools are not more useful because they fail to obtain key information needed to decide if the intervention is ready for the market. Similarly, inclusion of women is only "unduly" costly when the cost does not significantly augment the value of the data. Here again, the marginal value of including women is unknown but possibly very high (Report of the Planning Panel of the IOM
Division of Health Sciences Policy, "Issues in the Inclusion of Women in Clinical Trials," 1–2 March 1991).
Government policy does at this time limit the full participation of women in research trials. Specifically, Food and Drug Administration (FDA) policy restricts participation of fertile women in Phase I, Phase II, and large-scale Phase III trials, absent certain findings from optional animal research (FDA, "General Considerations for the Clinical Evaluation of Drugs," 1977). Further, Department of Health and Human Services (DHHS) regulations single out pregnant women, and require that research classes be therapeutic only, with minimal risk to the fetus (45 C.F.R. Pt. 46, Subpart B, "Additional Protections Pertaining to Research, Development, and Related Activities Involving Fetuses, Pregnant Women, and Human In Vitro Fertilization").
The exclusion of fertile women from even Phase IV and small-scale Phase III trials, however, as well as the frequent exclusion of all women from Phase I and II trials, testifies to the overbreadth of application of these policies at the level of individual pharmaceutical companies or principal investigators. Further, even the DHHS regulations, which purport to limit the autonomy of pregnant women on behalf of state interests in fetal outcome, are themselves subject to constitutional scrutiny, as they appear to elevate concerns for fetal well-being over concerns for maternal and female well-being.
Although state interest in fetal well-being is certainly permissible (Planned Parenthood of Southeastern Pennsylvania vs. Casey, ), it is not superior to a woman's own health and autonomy interests (In Re A.C., 573 A.2d 1235 [D.C. Cir. 1990]) and can even be subordinated to an inchoate desire simply not to have genetic offspring, without implications of bodily autonomy (Davis v. Davis, 842 S.W.2d 588 ).
State interests in fetal well-being may increase as viability is achieved (Roe v. Wade, 410 U.S. 113 ; Planned Parenthood of Southeastern Pennsylvania v. Casey). But state interest in previable fetal life, and in preconception potential fetal life, is strictly limited. Even if excluding pregnant women intending to go to term from all nontherapeutic research can be justified on the grounds of fetal protection, excluding all fertile but not currently pregnant women certainly cannot.
4. Does the disparity in benefit operate to disadvantage a protected class of persons? If it does, do the above justifications pass muster?
The Fifth and Fourteenth Amendments to the U.S. Constitution, as mentioned above, require that government actors refrain from denying "equal protection" of the laws under most circumstances. Specifically, where neither the groups suffering the discrimination nor the nature of the discrimination implicates special rights or disadvantaged populations, the discrimination is tolerable as long as there is a rational basis for the government actors' policy. Where, however,
fundamental rights are implicated (e.g., the right to marry) or the policy has the purpose or effect of further burdening already disadvantaged classes, the policy is subject to "strict scrutiny" (United States v. Carolene Products, 304 U.S. 144 ). Classes of discriminated persons for whom the strict scrutiny standard has been applied include aliens, illegitimate children, and certain racial groups. The theory is that these groups, whether owing to de jure impediments (aliens), stigma (illegitimate children), or significant, structural political obstacles (racial minorities), are unable to fully employ the political system and the protections of majoritarian democracy to safeguard their interests. They must therefore be "protected" by the courts from the excesses of the polity.
When strict scrutiny is applied to a governmental policy having the purpose or effect of burdening the exercise of fundamental rights or discriminating against a "protected" class, courts will strike down the policy unless it can be shown that (1) the policy furthers a "compelling" governmental purpose; (2) the purpose is being achieved by the least restrictive means possible; and (3) the policy is narrowly tailored to its goal, neither over- nor underinclusive.
To the extent that excluding fertile women burdens the fundamental rights to life and liberty, these policies are subject to the strict scrutiny test. This is a test they cannot meet. The government's purpose—protection of fetal life and maximization of the cost effectiveness of research—may well be compelling. But the purpose is not being met by the least restrictive means possible; for example, protection of fetal well-being could be achieved by requiring animal studies on fetotoxicity to precede human studies, and contraception and pregnancy monitoring could be used in human studies.
Further, the policy is both under- and overinclusive. It is underinclusive because it fails to contemplate, let alone address, fetal effects due to paternal exposures. Even if these are predicted to be less common and perhaps less severe than those due to maternal exposure, their complete absence from the policies is unacceptable. It is overinclusive because it excludes many women who, although fertile, are not going to become pregnant, or if pregnant, are not going to carry to term.
If one argues that excluding fertile women does not burden the exercise of a fundamental right, there is still the question of whether the exclusion burdens a "protected" class. Gender discrimination has been held to affect a class—women—who, although historically disadvantaged, do not suffer from the same obstacles as those faced by other "protected" classes. In light of their experience, however, the Supreme Court has deemed them to occupy an "intermediate'' status, somewhat like ''semi-protected" (Reed v. Reed, 404 U.S. 71, 1971) The result is that government policies having the purpose or effect of discriminating against women are accorded an "intermediate" level of judicial review.
Functionally, the Supreme Court has interpreted this to mean that
"classifications based on gender must serve important governmental objectives and must be substantially related to the achievement of those objectives" (Craig v. Boren, 429 U.S. 190, 1976), or, as stated in a more recent case, the government must have an "exceedingly persuasive justification" (Mississippi University for Women v. Hogan, 458 U.S. 718, 1982). Under this analysis, the justifications for excluding fertile women from research are somewhat stronger, but still subject to attack as being ineffective guarantors of fetal well-being and so overinclusive as to fail the test of "substantial relationship" or ''exceeding'' persuasiveness. It is argued, however, that discrimination against fertile women is not discrimination against women per se. Thus, as fertile women have never achieved even an intermediate protected status in constitutional jurisprudence, the only standard of review applicable to these policies is that of a mere rational relationship between the government interest and its policy.
The question of whether discrimination based upon pregnancy or the capacity to become pregnant is tantamount to gender discrimination has occasioned significant hairsplitting controversy. If government action is gender-neutral on its face, e.g., if discrimination based on the capacity to become pregnant is gender-neutral because some women, as well as all men, are unable to become pregnant, then despite the discriminatory results of a policy's application, a constitutional violation is demonstrated only if the discrimination is intentional (Personnel Administrator of Massachusetts v. Feeney, 442 U.S. 256, 1979).
In the context of employment law, and owing in part to the existence of Title VII's special provisions guaranteeing that pregnant women be treated like nonpregnant women for employment purposes (42 U.S.C. Sec. 2000e(k)), discrimination based on fertility has been held to be equivalent to discrimination based on gender (U.A.W. v. Johnson Controls, 111 S. Ct. 1196, 1991). But even compensated service as a research subject would not appear to meet the definition of "employee" under Title VII.
In nonemployment contexts, discrimination based on fertility or pregnancy has frequently been regarded as something distinctly different from discrimination based on gender, despite its obviously disparate impact on male and female populations (Geduldig v. Aiello, 417 U.S. 484, 1974; Michael M. v. Superior Court, 450 U.S. 464, 1981; Toomey v. Clark, 876 F.2d 1433, 9th Cir. 1989; U.S. v. Flores, 540 F. 2d 432, 9th Cir. 1976). Under this analysis, once again only a rational relationship must be demonstrated between the discrimination and the government purpose, unless the discrimination is shown to be intentional.
To this author, it appears that the exclusion of fertile women from research protocols burdens the fundamental liberty and right to life of fertile women, and functionally burdens the right to life of all women, who are thereby denied the benefits of research on women prior to marketing new drugs and interventions. Thus, a strict scrutiny of the policy would reveal that it is unconstitutional. Alternatively, an argument can be made that discrimination against fertile women, in this context, constitutes discrimination against all women, and thereby must be tested against an intermediate standard of review. Thus, in the context of government controlled research, the Equal Protection Amendment would appear to preclude exclusionary policies because they unconstitutionally deny access to a possibly important benefit to many and possibly most women. Research funded entirely by private means is not similarly covered by the amendment, and even limited receipt of governmental funds may not be sufficient to transform those actions into the sort of "state actions" subject to the amendment (Rendall-Baker v. Kohn, 457 U.S. 830, 1982; Stanturf v. Sipes, 224 F. Supp. 883, W.D. Mo. 1963; Merton, 1993.) Therefore, this research may well be immune from constitutional challenge.
Cotton, P. 1990. Is there still too much extrapolation from data on middle-aged white men? Journal of the American Medical Association 263:1049.
Society for the Advancement of Women's Health Research (SAWHR) 1991. Toward a women's health research agenda: Findings of the scientific advisory committee. Washington, D.C.: SAWHR.
Levine, R. 1988. Ethics and Regulation of Clinical Research, 2nd ed. New Haven: Yale University Press.
Levi, J. 1991. Unproven AIDS therapies: The Food and Drug Administration and ddI. In: Biomedical Politics, K. Hanna, ed. Washington, D.C.: National Academy Press.