Implementation of Web-Based Respondent-Driven Sampling to Recruit Users of Electronic Nicotine Delivery Systems in Brazil: Cross-Sectional Survey.

backgroundThe marketing of electronic nicotine delivery systems (ENDSs) has been prohibited in Brazil since 2009, and their regular use is less prevalent than in countries where these devices are not banned. To monitor the presence of ENDSs, it is important to prevent the development of a new generation of nicotine-dependent individuals. However, traditional surveys are costly for accessing rare populations. Therefore, to reach ENDS users aged ≥15 years, we used the online version of the respondent-driven sampling method (web-RDS), a peer chain recruitment method for contacting hard-to-reach groups.

objectiveThis paper aims to provide information on the implementation of the first web-RDS study in Brazil to recruit ENDS users.

methodsThis study was conducted in Rio de Janeiro, the second largest city in Brazil. After a formative phase using qualitative in-depth interviews, we selected the first participants ("seeds") to complete an online quantitative questionnaire on the profile of their own ENDS use and the size of their contact network of ENDS users. Participants received 3 coupons to invite eligible peers. For participation and recruitment, each participant received a gift card worth approximately US $4. The target sample size was 300 ENDS users based on a conservative estimate and adjusted for design effect.

resultsFrom August 2022 to May 2023, of the 12 seeds included, 508 attempts at access were recorded in the data collection system, of which 330 (65%) were eligible. Duplicate or ineligible attempts were identified and removed through automated and manual checks. Recruitment was initially slow due to the low monetary incentive, but it improved after the amount was increased. We found that 43.1% (75/174) of recruiters recruited only 1 eligible participant, 34.5% (60/174) recruited 2 eligible participants, and 22.4% (39/174) recruited 3 participants. Web-RDS was able to reach individuals in different areas of the city. Convergence was reached for target variables (ie, age and age at first use of electronic cigarettes). The median time to complete the questionnaire was 12 (IQR 8-17) minutes. Half (154/324, 47.5%) of the respondents reported that they knew up to 5 other ENDS users.

conclusionsThe web-RDS methodology proved to be a feasible approach for accessing the population of ENDS users in Brazil. Incentives for participation and recruitment emerged as a determining factor in the data collection process. However, researchers needed to be aware of individuals attempting to circumvent the system by participating without being eligible or participating multiple times. Implications for optimizing web-RDS are discussed. On the basis of the method's performance in this study, web-RDS shows potential to support future repeated data collection processes that could help monitor changes in the profiles of ENDS users over time, supporting the implementation of ongoing measures from Brazil's National Tobacco Control Policy.