Choosing drug use measures for treatment outcome studies. II. Timing baseline and follow-up measurement.

The use of different approaches to measurement in drug abuse treatment outcome studies has resulted in a lack of comparability across studies. This paper reviews different approaches to timing of baseline and follow-up periods and to dealing with time periods during which follow-up subjects are not "at risk" for drug use. Length and timing of baseline and follow-up periods are considered as well as periods of time within which specific drug use outcomes are measured. It is suggested that research focus on describing the natural course of drug use both prior to and following treatment, in order to determine the most appropriate length and timing of follow-up periods. It is recommended that investigators report drug use data from both "at risk" and "not at risk" periods, and that they choose methods of controlling for time "at risk" which do not eliminate important drug use data from analyses.