Factors Associated With use of HIV Prevention and Health Care Among Transgender Women

Seven Urban Areas, 2019-2020

Kathryn Lee, MPH; Lindsay Trujillo, MPH; Evelyn Olansky, MPH; Taylor Robbins, MPH; Christine Agnew-Brune, PhD; Elana Morris, MPH; Teresa Finlayson, PhD; Dafna Kanny, PhD; Cyprian Wejnert, PhD


Morbidity and Mortality Weekly Report. 2022;71(20):673-679. 

In This Article

Abstract and Introduction


Transgender women* are disproportionately affected by HIV. Among 1,608 transgender women who participated in CDC's National HIV Behavioral Surveillance (NHBS) during 2019–2020, 42% received a positive HIV test result.[1] This report provides results from seven U.S. urban areas where the 2019–2020 NHBS questionnaire was administered. Thirty-eight percent of participants reported having previously received a positive test result for HIV. Detrimental socioeconomic factors, including low income (44%), homelessness (39%), and severe food insecurity in the past 12 months (40%), were common and associated with lower receipt of HIV prevention and treatment services. Having a usual health care source or a provider with whom the participant was comfortable discussing gender-related health issues was associated with improved HIV prevention and treatment outcomes, including HIV testing, preexposure prophylaxis (PrEP) use, and viral suppression. These findings illustrate the benefit of gender-affirming approaches used by health care providers,[2] and highlight the challenging socioeconomic conditions faced by many transgender women. Ensuring access to gender-affirming health care approaches and addressing the socioeconomic challenges of many transgender women could improve access to and use of HIV prevention and care in this population and will help achieve the goals of the Ending the HIV Epidemic in the United States initiative.[3]

Initiated in 2003, NHBS conducts biobehavioral surveillance among persons at high risk for HIV infection. During June 2019–February 2020, NHBS surveyed 1,608 transgender women in seven U.S. urban areas using respondent-driven sampling. Eligible participants§ completed an interviewer-administered questionnaire and were offered an HIV test. The questionnaire included measures of gender identity, income, health insurance, housing,** food insecurity,†† HIV status, viral suppression (if HIV-positive), comfort with their health care provider in discussing gender-related health issues (hereafter referred to as comfort with a provider), unmet need for health care,§§ and usual source of health care. Because of racial and ethnic disparities in HIV prevalence, recruitment was focused on Black or African American and Hispanic or Latina transgender women as initial sampling recruits. Incentives were provided for completion of the interview and HIV test. Adjusted prevalence ratios (aPRs) and 95% CIs for prevention and treatment outcomes, by self-reported HIV status, were estimated using log-linked Poisson regression models with generalized estimating equations clustered on recruitment chain and urban area; models were adjusted for age, race and ethnicity, and urban area. Analyses were conducted using SAS software (version 9.4; SAS Institute). This activity was reviewed by CDC and was conducted consistent with applicable federal law and CDC policy.¶¶

Data from 1,608 transgender women were included in this analysis (Table 1). Thirty-eight percent reported having previously received a positive HIV test result.*** Forty-four percent earned <$10,000 annually. During the past 12 months 39% experienced homelessness, and 40% experienced severe food insecurity. Nearly one third (31%) of participants were interviewed in Los Angeles. By urban area, reports of homelessness ranged from 22% to 59%, and reports of recent severe food insecurity ranged from 28% to 47%. Comfort with a provider varied by urban area from 66% to 91%.

Socioeconomic status and health care accessibility were associated with health outcomes (Table 2). Among participants who reported a previous positive test result for HIV, self-reported viral suppression was less common among participants who reported experiencing homelessness during the past 12 months (aPR = 0.88; p = 0.003), and the likelihood of viral suppression decreased as the number of nights of homelessness increased. Severe food insecurity (aPR = 0.84; p<0.001) and unmet need for health care (aPR = 0.89; p = 0.027) were also less common among participants who reported viral suppression. Comfort with a provider (aPR = 1.17; p = 0.007) was more common among participants who reported viral suppression. Similar associations were found for current use of antiretroviral medication. Having a usual source of health care was also associated with current use of antiretroviral medication (aPR = 1.16; p = 0.015).

Among participants who did not report a previous positive test result for HIV, testing for HIV during the past 12 months was more likely among those who reported having a usual source of health care (aPR = 1.16; p<0.001) and comfort with a provider (aPR = 1.12; p = 0.004) (Table 3). PrEP use was more common among participants who reported having health insurance (aPR = 1.54; p<0.001), a usual source of health care (aPR = 2.54; p<0.001), and comfort with a provider (aPR = 1.79; p<0.001), and less likely among participants who reported an unmet need for health care (aPR = 0.82; p = 0.050). PrEP use was also more common among participants who had experienced severe food insecurity than those who had not (aPR = 1.23; p = 0.024).

*Persons who were assigned male sex at birth and who currently identify as women or transgender women.
Respondent-driven sampling is a methodology similar to snowball sampling and is often used when trying to sample hard-to-reach populations. The method relies on multiple waves of peer-to-peer recruitment to achieve the desired sample size. https://www.jstor.org/stable/10.1525/sp.2002.49.1.11?seq=1
§Eligible persons were those who were aged ≥18 years, had current residence in a participating urban area, had not previously participated in the current survey cycle, had ability to complete the survey in either English or Spanish, provided informed consent, and reported a gender identity of woman or transgender woman and were assigned male sex or intersex at birth.
Participants were asked to report their current gender identity from the following response options: woman, man, transgender woman, transgender man, or a gender not listed here. Participants were able to select more than one response option.
**Participants were asked if they had experienced homelessness during the past 12 months, including living on the street, in a shelter, in a single room occupancy hotel, or in a car. They were also asked to provide the number of nights during the past 12 months that they experienced homelessness.
††Severe food insecurity was defined as having not eaten for a whole day because there was not enough money for food at some point during the past 12 months.
§§Having an unmet need for care was defined as a "time when you needed medical care but didn't get it because you couldn't afford it" during the past 12 months.
¶¶45 C.F.R. part 46.102(l)(2), 21 C.F.R. part 56; 42 U.S.C. Sect. 241(d); 5 U.S.C. Sect. 552a; 44 U.S.C. Sect. 3501 et seq.
***Among participants, 38% self-reported living with HIV during the interview and were asked questions related to HIV treatment. During postinterview HIV testing, an additional 4% of participants received a positive HIV test result, for a total of 42% of participants who received a positive HIV test result (https://www.cdc.gov/hiv/library/reports/hiv-surveillance.html). Those who did not report living with HIV during the interview were not asked about HIV treatment.