Abstract
Large-scale studies of addiction treatment employ two stages: select providers, then select patients to follow-up. Nonresponse due to noncooperation of providers and problems of locating and recruiting patients may bias the results. We review selection and attrition biases in previous work and in four major United States treatment studies in the 1990s: DATOS (N = 10,100 clients. 96 units), NTIES (N = 6,593/71), SROS (N = 3,047/99), and CALDATA (N = 3,045/86). We develop a standard approach, break down sampling and attrition rates, and discuss differences in client, program, and methodology factors. We conclude with some methodological recommendations for future follow-up studies of addiction treatment. [Translations are provided in the International Abstracts Section of this issue.]
Key Words: