Methods: Semi-structured phone interviews were conducted with 14 PIs involved in HIV epidemiological, prevention and intervention research involving online recruitment and data collection. The sampling frame drew on the AIDS Clinical Trial Information Services (ACTIS), HIV Prevention Trials Network (HPTN), and Center for AIDS Research (CFAR). Interviews were analyzed using Thematic Analysis and yielding high inter-rater reliability (k >.88 across all themes).
Results: There was consensus that social media targeted advertising increased sample sizes at relatively low cost; although reaching ethnic minority populations at HIV risk remained a challenge. Online recruiting also led to large numbers of ineligible participants affecting research budgets and increasing screening burden. Restrictions on the use of sexual language and images relevant to sexual minority populations by social media companies and IRBs also limited advertising effectiveness. Online anonymity while increasing recruitment and decreasing the probability of social desirability bias, posed challenges to data integrity especially in identifying minors. PIs used a range of strategies to ensure data integrity with anonymous participants, including deduplication protocols and requiring participants to share contact information. The most controversial themes concerned responsibility and control of confidential information. Some PIs made special efforts to familiarize themselves with social media policies and to inform participants of informational risks, while others believed online research posed no more risk than daily internet use and thus protection against such risks was the participants’ responsibility. PIs had different experiences with university IRBs and technical support, but all voiced frustration on the limited control they had over social media policies and the absence of national guidelines on participant protections for online recruitment and data collection.
Conclusions: Online HIV research provides benefits and challenges in reaching marginalized populations. However, there is a lack of clarity involving the responsibility of different stakeholders to address informational risk. The absence of scientific consensus and national guidelines has led to strategies for protecting participant privacy and data integrity differing in effectiveness. The nature of that guidance is discussed.