Overcoming Recruitment Challenges

To the Editor: Recruiting for HIV prevention trials has been, and continues to be, a challenge. This is particularly true as more HIV prevention products, such as vaginal microbicides, move into phase 3 effectiveness trials, where large sample sizes are needed. In the past, cohort and feasibility studies have examined motivations1-3 and willingness to participate4,5 in HIV prevention studies. Another approach has been to document demographic and other predictors of study participation.6 Most of these studies have been in the context of vaccine trials and have not dealt with other types of HIV prevention trials. Additionally, few studies have looked at why individuals do not enroll in HIV prevention trials when approached by study recruiters (ie, those participants who never make it to the formal screening process). Documenting why individuals do not enroll in clinical studies is important for more effectively targeting willing and eligible populations from which to recruit. In this report, we present recruiting data from a randomized, placebo-controlled, safety and feasibility study of ACIDFORM gel (SRI International, Inc., Menlo Park, CA) with a diaphragm in South Africa. The study had a target enrollment of 120 women, which was successfully attained after 12 months of recruiting. Eligibility criteria included, among other things, women who were between the ages of 18 and 48 years, not pregnant or planning on getting pregnant, in a sexually active and mutually monogamous relationship, not currently breast-feeding, and HIV-negative. Trial participation required that participants use a cervical diaphragm and vaginal microbicide every time they had intercourse for the 6-month duration of the study. Recruiting for the study took place in community centers, public parks, and public health clinics within Johannesburg. All potential study participants were given a brief initial description of the study and eligibility criteria. Women who showed no interest at this point were asked the reason for not wanting to join, and the reason was noted. Women who expressed interest were invited to the study clinic for screening, and their contact information was obtained. Those women who went to the study clinic were given additional study information. During this time, if a woman refused screening or was ineligible for the study, her reason was noted. All failed recruiting efforts were documented with a recruitment log, which noted the date and time and reason for not screening. All reasons noted are based on participants' responses and were documented at face value (ie, not probed for further information). No names or identifying information was recorded in this log, nor was any demographic information recorded. None of the women in our sample underwent formal screening. Altogether, field staff approached 1144 women who did not screen for the study. In 2 cases, reasons were not properly documented and so were removed from the data set. The reasons were coded into 15 substantive categories. Reasons that did not comprise at least 0.5% of the total were placed into an “other” category. As illustrated in Table 1, the most common reasons for not screening were related to eligibility criteria, such as breast-feeding (20%), not sexually active (14.1%), and pregnant or planning on becoming pregnant (13.2%). Other commonly cited reasons were lack of interest (13.1%) and fear of HIV testing and/or receiving test results (10.6%). Only 3.5% of the reasons cited related to the study product, and to the diaphragm exclusively.TABLE 1: Reasons for Not Enrolling (n = 1142)Most of the reasons for not screening for the trial had to do with factors beyond the control of potential participants and were not related to the study product. Findings point to the importance of focusing recruitment efforts more precisely. Study recruiting materials, such as advertisements, could be more explicit in describing eligibility criteria. Formative research could also be useful by identifying areas where eligible participants are likely to be found in concentrated numbers. Sampling strategies that rely on participants' social networks, such as respondent-driven sampling,7 are designed to recruit individuals with similar characteristics and could be helpful in streamlining the recruitment process. Researchers might also consider, when possible, loosening or waiving eligibility criteria not essential to participant safety or the study design so as to accommodate a wider range of participants. This, however, was not an option for the study described; as a safety study, it was important to adhere strictly to the eligibility criteria so as to avoid unnecessary complications or confounding factors influencing the data. Recruitment efforts may also need to be adapted during the enrollment period. In the study described, the recruitment strategy changed over the course of the trial, as staff identified areas that did not yield good numbers. Recruitment catchment areas were geographically expanded into other communities. Initially, recruitment focused on communities directly surrounding the study clinic (eg, Yeoville Clinic [ie, Hillbrow, Berea, Yeoville, Bez Valley]). Efforts were then expanded to the Alexandra Township in hopes of finding women at lower risk for HIV so as to render a more favorable HIV-positive-to-HIV-negative ratio. Recruitment efforts were also shifted from public health clinics to HIV testing and counseling clinics to try and reduce the number of women who might decline based on concerns about HIV testing (eg, they had already tested HIV-negative). Additionally, community involvement proved to be invaluable in improving recruitment efforts throughout the study. Study staff attended community meetings, gave presentations to community organizations, and consulted the inner city community advisory group. The study also set up a referral system with a number of community organizations for ineligible participants, and a number of organizations referred potential participants to the study clinic. Additionally, the study involved experienced individuals from other projects to analyze the level of community engagement and advise community health workers with regard to their messaging about the gel and diaphragm. Experience from this trial (and others) has shown that planned and concerted efforts toward building rapport and trust between researchers and communities are well worth the time and effort. ACKNOWLEDGMENTS This research was funded by the United States Agency for International Development. Greg Guest, PhD* Lawrence Severy, PhD* Claire von Mollendorf, MD† Lut Van Damme, PhD‡ *Family Health International Durham, NC †University of Witwatersrand Johannesburg, South Africa ‡CONRAD Arlington, VA