Respondent Driven Sampling (RDS) in Asian American (AA) health research: Preliminary recruitment analysis from a community health survey in Chicago's Cambodian and Vietnamese neighborhoods

Background: Collection of representative health data from diverse AA populations remains a barrier to identifying community needs. The use of RDS in hidden populations is well established but not validated in recruiting members of minorities in urban settings.

Objective: To describe RDS in the context of recruiting participants for a population based health survey from Cambodian and Vietnamese communities in Chicago.

Methods: In September 2007, six initial recruits (seeds) were randomly selected through a telephone directory based on common surnames to initiate RDS. Of the six, two and four seeds were selected in Cambodian and Vietnamese communities respectively. Eligible participants were 18 years or older, self-identified ethnic Cambodian or Vietnamese, resided within the study’s geographical boundary, and spoke English, Khmer (Cambodian), Vietnamese, or Chinese.

Results: To date, four Vietnamese seeds remain in active recruitment status. Through December 2007, two Cambodian seeds recruited 148 eligible participants in 13 recruitment waves. The majority of Cambodian participants were women (64.7%), the median age was 48 years (range 18-84), and the mean social network size was 40 persons. Women (Χ²=4.3, p=0.04) and those 60 years or older (Χ²=6.0, p=0.05) were significantly more likely to be stronger recruiters (referred 2 or more peers).

Conclusion: Using RDS is a time efficient and cost effective method to recruit AAs for health survey research, particularly in communities where the AA population is prevalent, but not densely populated. Further research should explore whether the RDS method is appropriate to reach a representative sample of minority populations in urban settings.