Selective participation in a RDS survey among MSM in Ceará, Brazil

a qualitative and quantitative assessment

Abstract

Introduction: Respondent Driven Sampling (RDS) was used to conduct a biological and behavioral surveillance survey (BBSS) in Fortaleza, Brazil in 2005 among men who have sex with men (MSM). The study recruited many more MSM of lower social classes than Time Location Sampling and Snowball Sampling studies conducted in Fortaleza previously by the study team. Although poorer MSM are arguably more important for public health purposes, a surveillance method should provide information about all of the MSM population at risk. Objective: to explore reasons for low participation of higher social class MSM in the BBSS. Methods: RDS was used to recruit 406 MSM in Fortaleza, 2005. Data were analysed using the RDSAT to adjust for network size and recruitment patterns. A small-scale qualitative debriefing with eight higher social economic status (SES) MSM and staff from two Brazilian NGOs was also conducted to understand why they did not participate in the cross-sectional study. Results: of the 406 participants, more than half (56.1%) of MSM were less than 25 years old. Only 7.0% were of high SES. This paper found that the differential recruitment of higher social classes is due to: siting of the offices in a poor downtown area, insufficient incentives, NGOs associated with lower SES MSM, lack of solidarity among MSM, traditional class and wealth divides, fear of discrimination, and concerns about testing in general and the confidentiality of test results in particular. Because network links between high and low SES MSM appear to be minimal, the failure should not be attributed solely to RDS’s networking sampling methodology. Conclusion:operational, cultural, and socio-economic factors are barriers to the participation of high SES MSM. Strategies to enhance representativeness include additional formative research to explore the inclusiveness of networks, strategies to respond to the needs of higher SES MSM and encourage participation.