Criterion intervals for pretreatment drinking measures in treatment evaluation.

Abstract

Inpatient (24; mean age 44) and outpatient (24; mean age 39) men alcoholics were interviewed about the amount of alcohol they had consumed daily and the number of alcohol-related arrests, hospitalizations and stays in residential treatment facilities over the 360 days before treatment. To ascertain the validity of their self-reports, their reports of jail, hospital and residential stays were compared with official records and to determine the adequacy of a 30 day criterion interval as a measure of pretreatment drinking, the subjects'' self-reported daily drinking dispositions were compared across 4 pretreatment intervals: 0-30, 31-90, 91-180 and 181-360 days. Pearson product-moment correlation coefficients indicated a high degree of correspondence between subjects'' self-reports and official records. When self-reports and record data were discrepant, inpatients tended to overreport alcohol-related events while outpatients tended to underreport such events. A 30 day pretreatment interval was not representative of longer pretreatment intervals, especially for the more severely impaired inpatients. Significantly more days of heavy drinking were reported during the 30 day than during the 181-360 day interval when data were adjusted for opportunity to drink. It is suggested that data representing a 1 yr pretreatment interval be used in evaluating the effectiveness of alcoholism treatment.