Validation of the inventory of drinking situations (IDS) to individualize alcoholism therapy

Project PREDICT is a multicentric randomized controlled trial involving alcohol-dependent patients who receive acamprosate/ naltrexone/ placebo (40/40/20%) for 3 months following inpatient detoxification. Patients are seen biweekly for the first 6 months, then 4 more times every 3 months. The clinical methods match those of the project COMBINE. One of our attempts to find criteria predicting treatment response employs the IDS, i.e. a self-rating instrument that differentially evaluates excessive drinking during positive and negative emotional states. To analyze the relation between IDS data and treatment outcome, we developed a new one-factor model. During the first 2 years, 195 patients in 6 centers were randomized. During the active treatment phase (i.e. first 6 months), 25% were lost for follow-up, 39% relapsed and 36% remained abstinent. The IDS general (g-) factor scores were normally distributed with a broad variance of the mood states in which excessive drinking occurred, ranging from negative to positive. The overall retention and abstinence rates are comparable to other relapse prevention studies. Since the IDS g-factor scores show substantial variation and symmetric distribution, we expect that these data will enable us to test the hypothesis that patients with predominant relief craving respond better to acamprosate, while reward cravers do better with naltrexone. (Funded by BMBF grant 01EB0110).