Reliability and validity of the form 90 interview.

OBJECTIVE: Alcohol consumption is a central variable in substance abuse research and treatment. The study reports the psychometric characteristics of the German version of the Form 90 interview for the assessment of recent alcohol consumption.
METHOD: Reliability was evaluated in a test-retest study (7 days) with 30 consecutively admitted psychiatric inpatients with alcohol dependence. Validity of Form 90 was assessed with a second sample of 60 alcohol-dependent inpatients.
RESULTS: Form 90 demonstrated good to excellent retest reliability for the central variables of alcohol consumption. Retest reliability Pearson correlation coefficients (r) ranged from 0.76 to 0.99 and Intraclass Correlation Coefficients (ICC) ranged from 0.74 to 0.98. Reliability coefficients proved to be stable throughout the whole assessment window (90 days). Confidence intervals demonstrated sufficient precision of measures of alcohol consumption. Validity was supported by significant correlations of the amount of alcohol consumed with outcome parameters of the Lifetime Drinking History Interview (r=0.47 and 0.58), gamma-glutamyltransferase on admission to detoxification treatment (r=0.37), physical withdrawal symptoms (r=0.56) and the withdrawal symptom index (r=0.48).
CONCLUSIONS: Form 90 presented evidence of good to excellent psychometric properties, and its applicability to clinical populations of alcohol-dependent patients in German-speaking countries could be confirmed. Copyright (c) 2005 S. Karger AG, Basel.