Tracking substance users in a low income population: What does it take to attain high response rates?

Background: Extensive tracking using multiple contact methods is believed to generate higher response rates, especially in low-income populations, among people with substance use disorders, and in other hard-to-reach groups. However, the literature contains few precise descriptions of the methods used to track these populations in longitudinal studies, or of the effects of extended tracking on subject retention.

Methods: Hard-to-reach respondents who were successfully contacted only after extended tracking effort were identified using data and fieldwork tracking logs from the 12-month follow-up of 688 welfare recipients in a California county. An operational definition of extended tracking was developed based on interviews with a range of informants in U.S. survey research centers and from descriptions available in the literature. Descriptive comparisons were used to estimate the effects of extended tracking efforts.

Results: Over 30% of all respondents received at least one residential field visit, and almost 10% received at least one nonresidential visit. Extended tracking effort substantially increased response rates in the full sample (89% vs. 71%) and in respondents who self-reported problem drinking (81% vs. 63%), weekly drug use (87% vs. 66%), substance dependence (80% vs. 60%), and co-occurring substance dependence and mental health problems (55% vs. 36%). Over 35% of hard-to-reach respondents were interviewed in the field, compared to 5% of respondents who were found without extended tracking.

Conclusions: Extended, costly tracking efforts increase subject retention and may be justified in studies of substance-using and hard-to-reach populations.