JMIR Publications

ABSTRACT

Background:

The HIV epidemic disproportionately impacts transgender women in the United States. Cohort studies identify unique risks for affected populations, but use of facility-based methods may bias findings towards individuals living in research catchment areas, more engaged in health services, and/or, in the case of transgender populations, those who are open about their transgender identity. Digital clinical trials and other online research methods are increasingly common, providing opportunity to reach those not commonly engaged in research. Simultaneously, there is a need to understand potential biases associated with digital research, how these methods perform and whether they are accepted across populations.

Objective:

This study aims to assess the feasibility of developing and implementing an online cohort of transgender women to assess risks for HIV acquisition and other health experiences and to evaluate how an online cohort compares to a site-based, technology enhanced cohort for epidemiologic research. The overarching goal is to estimate incidence of HIV and other health outcomes among transgender women in eastern and southern United States.

Methods:

This sub-study is part of a larger multisite prospective cohort (LITE) conducted among transgender women, which also includes a site-based cohort of six eastern and southern US cities. The online cohort was launched to enroll and follow participants across 72 cities with similar demographic characteristics in the same regions as the site-based cohorts. Participants are followed for 24 months. Adult transgender women are recruited via convenience sampling (e.g., peer referrals, social media, and dating apps). Participants reporting negative or unknown HIV status are enrolled in a baseline study visit, complete a socio-behavioral survey, and provide oral fluid specimens to test for HIV. Participants not living with HIV (lab-confirmed) at baseline are offered enrollment into the cohort; follow-up assessments occur every six months.

Results:

Enrollment into the online cohort launched in January 2019. Active recruitment stopped in May 2019 and enrollment officially closed in August 2020. A total of 580 participants enrolled into and are followed in the cohort. A recruitment-enrollment cascade was observed across screening, consent, and completion of study activities. Implementation experiences with HIV test kits highlight the need for heavy staff engagement to support participant engagement, visit completion, and retention, even with automated digital procedures.

Conclusions:

This study is responsive to increasing research interest in online cohort research, particularly for populations who are most affected by the HIV epidemic and for those who may otherwise not participate in-person. The progression across stages of the recruitment-enrollment cascade provides useful insight for implementation of cohort studies in the online environment.