Hillary Ash explores the politics of gender and women in early AIDS epidemiology and data collection
On September 7th, 1993, the American Broadcasting Company’s (ABC) World News Tonight aired a segment on the growing number of women with AIDS in the United States. They featured the story of Beverly, a woman who repeatedly sought out physicians’ expertise to identify the cause of a strange set of long-term symptoms she had been experiencing. Eventually, one physician tested her for HIV, and Beverly’s test came back positive. Beverly explained to the ABC reporter that “[the doctors] never looked for this disease in me, being as how they thought it was a gay men’s disease” (Treichler and Warren 1998, 109).
Throughout the 1980s in their public health messaging, United States government leaders and biomedical experts stressed that anyone could be infected with HIV. However, the association between gay men and HIV/AIDS remained unshakeable in American culture even in healthcare settings. This bias towards an archetypical ‘AIDS patient’ influenced cases like Beverly’s to the extent that neither she nor her doctors considered testing for HIV despite the nation’s ever-increasing AIDS numbers.
The problem of women’s invisibility in the United States’ epidemic was not lost on AIDS activists at the time. While bias towards a certain kind of person with AIDS was no doubt multi-causal, many activists identified the Centers for Disease Control’s (CDC) epidemiological surveillance data collection as flawed, especially with regards to sex-based AIDS data.
Due to privacy concerns and HIV stigma, the CDC only collected data on AIDS. (HIV surveillance data would not be collected in the United States until 2008 due to fear that the de-identified forms would lead to duplicated cases and thus artificially inflate HIV morbidity and mortality numbers.) Data collection occurred through de-identified “Case Report Forms” sent to the CDC by local and state public health authorities, and authorities constructed these forms based on the CDC’s AIDS surveillance definition. An epidemiological surveillance definition can either be highly specific—meaning that a patient has to meet a high threshold of criteria in order to be counted as a case—or highly sensitive—the threshold is lower and thus more instances of misdiagnosis slip into surveillance data. In the case of AIDS, public health authorities chose to utilize a highly specific definition that resulted in a narrow set of criteria a patient had to meet in order to be counted as an official AIDS case. Because AIDS was first diagnosed in gay men, much of the definitional criteria reflected AIDS as it manifested in gay men. The CDC did not include gynecological-specific infections or diseases in their AIDS surveillance definition until 1993 despite widespread anecdotal evidence suggesting persistent gynecological infections correlated with HIV infection. Thus, it was harder for women to be seen by the surveillance system.
In addition to definitional problems, the hierarchy of risk used by the CDC in their data collection to identify likely exposure routes contributed to women’s erasure in the epidemic. In this hierarchy, the first category of exposure risk always identified homosexual or bisexual men as the number one transmission category for HIV. This category emphasized the identity homosexual or bisexual as a risk rather than any specific sexual behavior until 1989. The second risk category was intravenous drug use (IVDU). In the early 1980s, only one route of exposure could be marked on the “Case Report Form,” so if a man with AIDS was both homosexual/bisexual and an intravenous drug user, his mode of AIDS transmission would only be identified as homosexual/bisexual male. Later, multiple modes of exposure would be introduced in surveillance reporting.
The use of hierarchical transmission rounds was significant for women in two ways. First, for much of the 1980s, multiple modes of AIDS exposure did not exist for women, only men. Even if women met the high threshold of AIDS diagnosis, there was far less nuance to the exposure data about them. Second, with male homosexuality/bisexuality holding the highest position in the single-cause hierarchy, sexuality—and by extension sex—was over-represented as a risk category. Because of how the hierarchy functioned, homosexual or bisexual men who were exposed to HIV via drug use were counted in the surveillance data as having sex-based exposure. Cindy Patton (1994, 97) noted that when the CDC introduced multiple risk categories, cases of “Homosexual/Bisexual Men” as the definitive exposure route dropped to 50%. Had IVDU—a category that women could occupy—been at the top of the exposure hierarchy from the start, “Homosexual/Bisexual Men” would have likely only ever represented 50% of the cases, leaving space for AIDS to be perceived as a condition women could also have.
Despite its hyper-focus on sexuality as the primary route for HIV transmission, the CDC did not account for a spectrum of sexualities in women in their Weekly Surveillance Report, a document that provided AIDS morbidity and mortality statistics to physicians and the public. Women could only fall into the following transmission categories: IVDU, Haitian, Hemophilia, Heterosexual Contact, Transfusions with Blood/Blood Products, or None of the Above/Other. For nearly the entire epidemic, the CDC’s surveillance system had no category for women who were exposed via homosexual or bisexual contact. The exception was a June 15, 1982 CDC report that included categories for both “Heterosexual Females” and “Bisexual Females” with one case of AIDS reported in a bisexual woman (Centers 1982a). By the publication of the August 6, 1982 report, the categories “Heterosexual Females” and “Bisexual Females” were replaced with the single category “Females” (Centers 1982b). As an unmarked term, “females” should be understood as implicitly communicating “heterosexual females.” The bisexual woman with AIDS who had appeared in the CDC’s data not even two months before had become heterosexual. It is possible that the CDC relegated cases of AIDS in homosexual and bisexual women into the “Other” categories, which would later be included. Women appeared in greater numbers in catch-all categories such as “undetermined mode of exposure” or “no identified risk” than men, categories which would often be the second or third largest transmission category for women with AIDS.
While there were significant problems with data collection about women generally, data on women of color with AIDS was not publicly circulated in the Reports until 1989 when reporting shifted from weekly to monthly. Prior to 1989, the CDC reported morbidity and mortality data based on sex and race/ethnicity separately. A reader could find statistics on AIDS rates in African Americans or women but not African American women. If they hoped to find any data about “Asian/Pacific Islanders” or “American Indian/Alaskan Natives,” readers might have to look in the “Other/Unknown” column or footnotes depending on the year (Centers 1988).
Exposing flaws in epidemiological surveillance infrastructures in one way to reveal how bias unintentionally becomes woven into what is culturally lauded as “objective” science. Once hidden behind the veil of objectivity, medical injustices are perpetuated against those communities who have a long history of fraught tensions with medicine. Countless women died from AIDS. While they may have never officially counted, we must lift the shroud of silence around the imperfect system that allowed their deaths to go unheard.
About the author
Hillary A. Ash is a Visiting Lecturer at the University of Pittsburgh researching rhetorics surrounding women and medical injustices. She completed her PhD on women’s invisibility during the first decade of the U.S. AIDS epidemic in 2020.
Centers for Disease Control. 1982a. Kaposi’s Sarcoma (KS), Pneumocystis Carinii Pneumonia (PCP) and Other Opportunistic Infections (OI): Cases Reported to the CDC, as of June 15, 1982.
Centers for Disease Control. 1982b. Kaposi’s Sarcoma (KS), Pneumocystis Carinii Pneumonia (PCP), and Other Opportunistic Infections (OI): Cases Reported to the CDC as of August 6, 1982.
Centers for Disease Control. 1988. AIDS Weekly Surveillance Report – January 4, 1988.
Patton, Cindy. 1994. Last Served?: Gendering the HIV Pandemic. New York: Taylor & Francis.
Treichler, Paula and Catherine Warren. 1998. “Maybe Next Year: Feminist Silence and the AIDS Epidemic,” in Gendered Epidemic: Representations of Women in the Age of AIDS, ed. Nancy L. and Katie Hogan Roth. New York: Routledge.