Predictors of attrition in a longitudinal study of substance abusers.

Previous research has emphasized the importance of minimizing attrition in longitudinal studies. The authors examined the influence of demographic, clinical, and process factors on attrition from a longitudinal study of 286 substance abusers recruited at a central intake unit. Univariate tests showed that those who completed three, six, and 12 month interviews had higher baseline alcohol and drug use and were more likely to provide three or more contacts when recruited, to be female, to have been married, and to have previously received substance abuse or psychiatric treatment. Multivariate logistic regression revealed that complete data participants (those who completed all three ASI interviews) were nearly three times more likely to have previously received psychiatric treatment and twice as likely to have completed an early engagement interview. In addition, they were twice as likely to be female, and nearly twice as likely to be employed, own a residence, have completed high school, and be married. Procedures that may help decrease study attrition are discussed.