Heterogeneity and Social Difference
Having campaigned against narrow inclusion criteria, activists pushed the question of access to trials even further: they opened up the issue of the social demographics of clinical trial participants. In public, the research establishment was on weak ground in this debate: since AIDS activists had successfully promoted the notion that access to trials was potentially beneficial, ordinary notions of equality and justice suggested that this social good should be distributed widely and fairly. Just the opposite was in fact the case, at least in the trials funded through NIAID's ACTG program.
As a 1989 front-page Los Angeles Times exposé by Robert Steinbrook had revealed, "blacks, Latinos and intravenous drug users, the groups increasingly afflicted with AIDS virus infections, are significantly under-represented in federally sponsored AIDS clinical trials…." Using documents obtained from NIAID under the Freedom of Information Act, Steinbrook showed that while blacks and Latinos accounted for 42 percent of adult U.S. AIDS patients, they made up only 20 percent of the research subjects in the ACTG trials. Only II percent of the ACTG subjects were injection drug users, though this population accounted for 28 percent of AIDS cases. A later study by New York activists that was presented at the Seventh International Conference on AIDS in 1991 showed that women comprised only 6.7
percent of ACTG trial participants. While women accounted for only 9.8 percent of overall AIDS cases according to the CDC's statistics, activists argued that many actual cases of women with AIDS were not captured by the CDC's surveillance definition.
In fact, most trials were populated largely by adult, white gay men who had contracted HIV through sexual transmission. Other demographics groups, such as men of color and women and children of all races, were underrepresented; so were those who had contracted HIV through drug use or contaminated blood products. The reasons for these exclusions were multiple. Certainly it didn't help that there were no AIDS Clinical Trials Units in five of the thirteen cities that contributed the greatest numbers to the CDC's statistics for U.S. AIDS cases—Houston, Philadelphia, Atlanta, Dallas, and San Juan, Puerto Rico. Injection drug users were rarely targeted for recruitment because they were considered to be "bad" research subjects. "They are alienated, disorganized and distrustful," one doctor told the New York Times in 1988. "They don't keep appointments; you can speculate about why, but they just don't do it." (In response, other doctors maintained that those drug users who were participating in drug treatment programs made perfectly good research subjects.) Some observers also cited "real-life barriers" to the participation of working-class and poor people in trials, including the cost of transportation and the lack of child care.
In the case of minority groups, particularly African-Americans, it also could not be assumed that people were anxious to volunteer for trials. Whites may have been banging on the doors demanding to be put to use as "guinea pigs," but many blacks had vivid historical memories of precisely what such use entailed. As one Dallas health educator testified before the National Commission on AIDS, "So many African American people that I work with do not trust hospitals or any of the other community health care service providers because of that Tuskegee Experiment." The fact that many researchers had targeted Africa as the site of origin of AIDS aroused further distrust in African-American communities. In one survey of black church members, 35 percent of the respondents expressed agreement with the claim that AIDS might actually be a form of genocide perpetrated by the U.S. government on minority communities. Given such sentiment, it was unlikely that the AIDS Clinical Trials Units would draw many African-Americans simply by posting an announcement and waiting for the phone to ring. At a minimum, serious recruitment
would have required a concerted outreach effort that presupposed gaining the trust of community leaders.
Finally, the very same reforms in the protection of human subjects that incidents such as the Tuskegee study had engendered also created pressures to exclude various groups from clinical research. According to this logic, experimentation was something that vulnerable populations were to be protected from . Children, for example, were generally not to be the subjects of clinical trials until after a drug had been proven safe and effective in adults. (Partly as a result, AZT was not widely distributed to children with AIDS in the United States until October 1989, more than three years after adults had gained broad access to the drug. )
The situation with women was even more restrictive. Until 1986, women "of childbearing age"—regardless of whether they were pregnant or had any intention of becoming so—were barred from trials as a matter of course out of fear of causing damage to fetuses (or out of fear of resulting lawsuits). Though the NIH formally changed its rules in response to protests, research protocols (for AIDS and other illnesses) often continued to exclude women, and the FDA continued to ban women from early research on new drugs. Some women with AIDS charged that even after they had offered to undergo sterilization, they were still told they would be unable to join the clinical trials for drugs that the women considered to be promising.
In practice, such policies meant not only that women lacked access to experimental therapies that might help them but also that doctors lacked certainty about the effects of drugs in women's bodies. "American women have been put at risk," said Representative Pat Schroeder, citing a study begun in 1981 that sought to investigate the role of aspirin use in preventing heart attacks, which had enrolled twenty-two thousand male doctors. "[NIH] officials told us women were not included in this study because to do so would have increased the cost," commented a congressional investigator. "However, we now have the dilemma of not knowing whether this preventive strategy would help women, harm them, or have no effect." Another female member of Congress, Olympia Snowe, described a federally funded study on the relation between obesity and cancer of the breast and uterus; the pilot study had used only men. "Somehow I find it hard to imagine that the male-dominated medical community would tolerate a study of prostate cancer that used only women as research subjects," Snowe commented.
Men's and women's bodies are manifestly different, and it seemed not unreasonable to suggest that drugs might have different effects in the different genders. Some experts claimed that the same was true for members of different racial groups; Asians, for example, were said to metabolize an antidepressant drug called desipramine more slowly. As to whether such differences might reveal themselves with AIDS therapies, researchers remained uncertain. Anthony Fauci would comment in 1994 that such a basic question as "whether the drug has an antiviral effect" is something "you could determine … by giving it to people who live on … Park Avenue and 69th Street in New York City and not worry about the rest of the world…." But that's different, Fauci went on to say, from the question of precisely how best to use particular drugs in particular populations—an issue to be determined by administering drugs to a heterogeneous collection of patients.
Whether or not there truly are significant variations in how drugs affect different social groups, the fact remains that the existence of such variations is widely considered plausible by a society that increasingly tends to understand social difference as something rooted in biology. The body of a gay white man, therefore, was seen as an inappropriate location for generating predictions about the effects of a drug on a straight white man, or a gay Latino man, or a gay white woman. The complexities involved in such debates became evident in 1991, when yet another AZT study, conducted in Veterans Administration hospitals by the Department of Veterans Affairs, purported to find a difference in response to AZT between white and minority patients. "AIDS Study Suggests Drug May Have Racial Limits," declared a Washington Post headline. Dr. Wayne Greaves, an infectious disease specialist at Howard University in Washington, D.C., told the New York Times that he would "counsel black and Hispanic patients that in light of the new data they had to decide for themselves whether to take the drug." But as scientists and activists scrutinized the data, they discovered that the Veterans Administration researchers had in fact made their claims about a racial category they called "minority." Lacking sufficient numbers of Latinos and African-Americans to draw statistically significant conclusions, the investigators had simply decided to lump the two groups together. Ron Johnson of New York City's Minority Task Force on AIDS blasted the "sloppy" methodology: "Until they do a credible study, they're just playing with us, throwing out confusing and conflicting bits of information." Underlying this response, however, was the prior reification of racial
categories: "black" and "Hispanic" were considered to be "real" markers of biological difference (and thus potential predictors of differences in response to treatment), in a way that the hybrid category "minority" was not.
The issue was indeed one of credibility, as Johnson indicated, and in U.S. society at the time, a drug would simply not be perceived as credible across the board if it had not been tested in a diverse range of social groups. As Vivian Pinn-Wiggins, a pathologist at Howard University and the president of the National Medical Association (an organization mainly of African-American physicians) put it in 1990, "some of our physicians are a little leery" of certain medications because "we can't be certain whether minorities have been participants" in the clinical trials. Of course, as Ellen Cooper pointed out at the "Methodological Issues" conference, there are many kinds of heterogeneity, and there are no a priori grounds for singling out particular instances, such as racial and gender differences, and assuming that there are the ones that will manifest differences in response to treatments. These categories are simply the ones with greater social and political salience.
Two sets of issues came together in the debate over homogeneity and heterogeneity in a study population: the need for a morally credible policy promoting fair access to experimental drugs and the need for a scientifically credible policy for acquiring generalizable data. Between these two sets of issues, AIDS treatment activists had plenty of room to play. Defenders of the notion that a "clean" trial required a homogeneous research population, by contrast, found themselves increasingly on the defensive. The activist critique demonstrated the back-and-forth movement between ethical and epistemological claims-making that AIDS treatment activism had perfected: heterogeneous trials were not only fairer, they were also better science. Though it would not always prove so easy, in this case, the goals of "access" and "answers" could be made to coincide.