Difficult-to-recruit respondents and their effect on prevalence estimates in an epidemiologic survey.

Generous contact efforts were used to recruit 3,004 respondents into the first wave of the St. Louis Epidemiologic Catchment Area (ECA) project, a psychiatric epidemiologic study of the general population, conducted from 1981-1982. These efforts were analyzed to establish which sociodemographic characteristics or current psychiatric disorders were determinants of difficulty in recruitment. Being young, male, black, a nonrural resident, educated, and full-time employed were the demographic characteristics associated with increased contact efforts. Persons currently meeting criteria for an alcohol disorder required almost 20% more contact attempts than those without the disorder; this difficulty lay both in their being less available for an initial contact and refusing at a higher rate once contacted. An optimal recruitment effort cutoff point is provided, and a method is suggested for detecting that optimal point in the course of ongoing studies.