Women's Participation in Clinical Research: From Protectionism to Access
Tracy Johnson and Elizabeth Fee
Nearly a decade before women's exclusion from clinical research became an issue of public scrutiny and political debate, a 1981 article documented women's underrepresentation in drug trials. The article observed that conservative Food and Drug Administration (FDA) policies which discouraged the inclusion of women of ''childbearing potential'' in early drug trials led to their widespread exclusion in later, Phase III testing.1 That the article stirred so little response reflects the relative consensus at the time among researchers and federal policymakers vis a vis protectionist policies governing research on women.
In contrast, the findings of the General Accounting Office (GAO) in 1990 met with public outcry and congressional action in the form of the Women's Health Equity Act.2 This GAO report documented that the National Institutes of Health (NIH) had made "little progress" toward implementing its 1986 policy encouraging the inclusion of women as subjects in clinical research.3 Although NIH's records retention methods render any attempt to quantify the extent of women's exclusion difficult, press coverage has focused attention on several large-scale clinical trials with all-male study populations, e.g., the Physicians' Health Study4 and Multiple Risk Factor Intervention Trial (aptly abbreviated as MR. FIT).5 The mobilization of congressional women, women in the health professions, and women's advocacy organizations around this issue marks a shift in both public opinion and policy development.
CURRENT FEDERAL POLICY
Following the issuance of the 1990 GAO report, NIH promulgated a strengthened policy governing the award of federal research grants:
Applications for grants and cooperative agreements that involve human subjects are required to include minorities and both genders in study populations so that research findings can be of benefit to all persons at risk of the disease, disorder or condition under study.6
A 1990 NIH Policy Notice further specifies that the inclusion of women in study populations will be considered a matter of scientific and technical merit in peer review. Justifications for exclusion of either gender must be "compelling" and supported by "a strong scientific rationale."7 As a result of these recent NIH actions, other federal agencies with jurisdiction over human research are revisiting their overlapping, and sometimes contradictory, policies.
The FDA, which oversees pharmaceutical drug testing, is presently reviewing its policies in the area.* FDA guidelines promulgated in 1977 advised that women of "childbearing potential" participate in trials only after early Phase II (efficacy and safety) trials and (female) animal reproductive studies have been completed. The guidelines provide an exception for research on life-threatening diseases.8 For the later phase III studies, FDA lifts restrictions on women's participation and, since 1989, has advised gender (and other "subgroup") analysis of trial data.9
In addition, federal regulations enunciated in the Federal Policy for the Protection of Human Subjects (45 C.F.R. pt. 46) require that local institutional review boards (IRBs) ensure the ethical design of research subject to federal regulation (e.g., by the FDA) as well as research funded by the Department of Health and Human Services (DHHS). IRBs are required to assure that the "selection of subjects is equitable" and to "be particularly cognizant of the special problems of research involving vulnerable populations, such as . . . pregnant women."10 Subpart B (which applies only to research funded by DHHS) offers "additional safeguards'' for research involving fetuses, embryos, or pregnant women.11
Thus, at one extreme, federal (NIH) policy makes no mention of fertility
issues and dictates women's inclusion in clinical trials; at the other, federal (FDA) guidelines advise the exclusion of premenopausal women, at least during the early phases of testing. The federal regulations set forth at 45 C.F.R. pt. 46 take a middle course. They promulgate a "minimal risk" policy toward embryos/fetuses and urge explicit informed consent procedures for women, advising them of the "unforeseeable" risks of unintended pregnancy. It is small wonder that IRBs, the pharmaceutical industry, and researchers report confusion.12 Indeed, a 1992 GAO report attributes the "underrepresentation" (ambiguously defined) of women in drug trials, as well as the lack of gender analysis, largely to imprecision in FDA guidelines.13
Public outcry and tragedy have driven policy development in the area of research on human subjects. The original Food and Drugs Act of 1906, for example, required no safety testing prior to marketing and instead monitored drugs' strength and purity.14 Congress passed a revised Federal Food, Drug and Cosmetic Act in 1938 only after the Elixir Sulfanilamide tragedy, which resulted in 107 poisoning deaths. The strengthened law required industry sponsors to prove drug safety before market release.15
Public concern and sensitivity to ethical issues in research were greatly strengthened by the Nuremberg war crime trials during the late 1940s, which revealed Nazi atrocities in human experimentation perpetrated on Jews and other "undesirables." As a result, the Nuremberg Code promulgated standards for the protection of human research subjects.16
During the 1960s, thalidomide-induced birth defects in Europe spurred public and congressional interest in drug development procedures in the United States. In 1962, amendments to the Food, Drug and Cosmetic Act effectively institutionalized a drug approval process resembling the one in place today. Provisions included requirements that manufacturers demonstrate safety and efficacy; that FDA collect adverse reaction reports; and that drug advertising clearly indicate associated risks as well as benefits of drug therapy.17
In response to new regulations as well as the threat of litigation, many pharmaceutical companies adopted increasingly conservative postures with respect to the participation of premenopausal women in clinical drug trials. Several commentators have observed that such continuing conservatism is counterintuitive given that liability incurred in the thalidomide (and subsequent) cases was due to inadequate drug testing and ignored research results which indicated the potential for reproductive harm in women.18 Presumably, an all-male study population would not have shielded the manufacturers of thalidomide from liability claims of a drug marketed to women.
The DES and Dalkon Shield tragedies followed closely on the heels of the thalidomide disaster and coincided with the grass roots women's health movement of the 1960s and 1970s. As a result, early women's health advocates—organizations such as the Boston Women's Health Collective and the National Women's Health Network—framed the issue of women's health research as one of inappropriate inclusion (rather than exclusion).19 They focused, for example, on reports that women were administered DES without their knowledge as a part of an experimental protocol.20 Hence, these activists directed their initial efforts toward improved informed consent procedures and other protective measures for women as research participants and consumers.21
Public disclosure of the Tuskegee Syphilis Study during the 1970s prompted the codification of the protectionism already emergent among industry and consumer advocates. This observational study involved rural African American men with syphilis who remained in the study (untreated) well after treatment was widely available.22 Partly in response to Tuskegee, the 1974 National Research Act called for the establishment of a National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research to "identify ethical principles" and to "develop guidelines" for research involving human subjects.23 Further impetus for creation of the National Commission was the raging political debate over abortion, artificial reproductive technologies, and fetal tissue research.24
In 1973, The Supreme Court case Roe v. Wade legalized abortion and galvanized the small but impassioned antiabortion movement in the United States.25 Abortion politics affected not only contraceptive and abortifacient (e.g., RU486) research, but all research involving women of "childbearing potential." Indeed, Fletcher and Ryan have remarked on the irony that 45 C.F.R. pt. 46 grants embryos and fetuses a higher standard of protection ("minimal risk") than that accorded to children, who may in some cases participate in research involving "greater than minimal risk."26 Others have observed that federal guidelines display more concern for (potential) fetal health than women's health.
In short, protectionist policies with respect to women's participation in clinical trials emerged from a variety of political, social, and legal forces: public outrage to unethical human experimentation, liability fears, consumer activism, and abortion politics.
ADVENT OF INCLUSIONARY POLICIES
AIDS activism concerning access to unapproved AIDS treatment therapies during the 1980s brought the first real challenge to protectionist policies.27 The National Commission's Belmont Report (1979) had anticipated the events to come, noting that:
Beneficence thus requires that we protect against risk of harm to subjects and also that we be concerned about the loss of substantial benefits that might be gained from research.28 [Emphasis added.]
In May 1987, the FDA responded to pressure from AIDS activists and issued regulations expanding access to experimental drugs used to treat "serious" and "life-threatening" illnesses.29 Activists also lobbied loud and hard for greater public accountability with respect to awarding federal research dollars. Increased federal appropriations for AIDS research testifies to their persuasiveness.
The women's community observed the success of AIDS activists in reorienting financial and scientific resources to better address AIDS research needs. AIDS activism provided women with a language and a political strategy with which to pursue a women's health research agenda. Changing demographics provided the critical mass. Women of the baby boom generation were now of an age to be increasingly concerned about the lack of attention to their health and well-being. Dramatic increases in medical school enrollment among women during the 1970s30 provided a vocal minority of medical professionals in the subsequent decade who questioned current priorities and policies in women health research.
Women observed that historically, medical and public health professionals have conflated women's health with reproductive and child health. Women's health needs have been subsumed under obstetrics and gynecology or addressed through maternal and child health programs. Neither the medical nor the health community perceived cardiovascular disease, for example, as "women's health," despite the fact that it is the leading cause of death for women as well as men.31 As a result, research in nonreproductive areas seldom explored gender-linked hypotheses and often excluded women altogether.
However, pressure from female researchers within NIH resulted in a two-year U.S. Public Health Service Task Force on Women's Health Issues, which concluded in its 1985 report that:
The historical lack of research focus on women's health concerns has compromised the quality of health information available to women as well as the health care they receive.32
Indeed, it was the recommendations of this task force which inspired the original NIH policy to encourage the inclusion of women in clinical trials.
Moreover, it was women like these—from within the medical and health professions—who ultimately convinced Reps. Pat Schroeder (D-CO), Olympia Snowe (R-MA), and Henry Waxman (D-CA) to request the influential 1990 GAO report. The ensuing public outrage—particularly among women—signaled a significant shift in popular opinion. Practices and policies once presented as
protective were now labeled paternalistic and discriminatory. As 1992 campaign rhetoric reminded us, these changes in perception occurred within a context of heightened gender resentment—at, for example, continued gender-based pay inequities in the work place, the erosion of abortion rights, and the Clarence Thomas-Anita Hill debacle.
Federal policies which address women's participation in clinical research were now scrutinized for bias. For instance, women's health advocates pointed to a curious section of 45 C.F.R. 46, which reads:
When some or all of the subjects are likely to be vulnerable to coercion or undue influence, such as children, prisoners, pregnant women, mentally disabled persons, and economically or educationally disadvantaged persons, additional safeguards have been included in the study to protect the rights and welfare of these subjects.33 [Emphasis added.]
One wonders what aspect of pregnancy renders women particularly vulnerable to "coercion" or "undue influence."
Women's health advocates have also observed the asymmetry in the risk/benefit analysis for research impacting male versus female reproductive potential. According to FDA guidelines, even research which has resulted in reproductive harm in male animals may go forward depending "upon the nature of the abnormalities, the dosage at which they occurred, the disease being treated, the importance of the drug, and the duration of drug administration."34
In contrast, a sweeping definition of "women of childbearing potential" by FDA virtually excludes all menstruating women from early testing of drug therapies. Moreover, language which implies that women "in certain institutions, e.g., prisons" might be acceptable study participants is remarkable given an ignominious history of research on institutionalized populations.35 Critics of FDA policy have questioned whether current policy is more a reflection of gender stereotypes—female susceptibility and male invulnerability—than of good science.36
Since the GAO report focused on NIH policy, NIH bore the brunt of the public censure after its publication. In response to criticism, NIH officials buttressed traditional justifications for excluding women—fetal protection and liability concerns—with an argument of cost.37 The inclusion of women in clinical research means that researchers must consider hormonal cycling and deal with larger sample sizes and recruitment "difficulties"—all of which, they argued, increase study costs.
These responses assume that the normal hormonal variations women experience throughout their menstrual cycles complicate research design by introducing an additional variable for which the researcher must control.
However, others counter that, for women, menstruation is as normal as breathing. Thus, to characterize this normal body process as "an extra variable" assumes that the male body is the norm. Critics do not dispute the existence of metabolic differences between the genders. Rather, they argue that such differences are precisely the reason that research must include both men and women.38
The hormonal cycling version of the cost argument fails to account for women's absence from behavioral and epidemiologic research. In these contexts, the cost argument reduces to one of sample size. Defenders of current practice argue that to represent all possible gender/racial/ethnic/age/class combinations in study populations would introduce onerous costs.39 The counterargument holds that the goal is not to mandate the inclusion of women in each and every research activity. Instead, federal research institutions should seek to ensure that their research portfolios, when viewed in their entirety, reflect the diversity found in the U.S. population—who, after all, underwrite the studies.40
Certain members of the research community maintain that women are more difficult to recruit into studies.41 While corroborating empirical evidence is often lacking, the mere perception of difficulty may itself be an obstacle. Women, they assert, require "special arrangements" such as child care and transportation. However, women's health advocates are quick to respond that women demonstrate greater health-seeking behavior than men and are generally more compliant patients and research participants.42 On balance, perhaps the observed behavioral differences between men and women merely point to the need for gender-specific strategies to recruit and retain both genders in clinical studies, rather than pose inherent obstacles to women's inclusion.
Ultimately, the cost argument falls apart when one considers the cost of the present gaps in information about women's health. Clearly, in the absence of data on female populations, women and their health providers will continue to apply to women the research results from studies on men. This approach minimizes neither risks to women nor costs to society. Nor does it protect future children from prenatal damage.
Much as the 1970s' women's liberation movement incorporated themes of the civil rights movement, women's health advocates of the 1990s borrowed the concept of a consumer-driven research agenda from AIDS activists. Both movements have focused constructive attention on the research process. However, the challenge remains to define a policy which will minimize risks and yet allow productive research for men, women, and children. It is clear that the protectionist tradition will no longer continue unchallenged, and that women are demanding input in the evaluation of relative risks and benefits inherent in research activities. Beyond mere inclusion, women insist that their diverse health
concerns should inform the formulation of the hypotheses, design, and implementation of federally funded research activities.