Predicting relapse and rehospitalization in alcohol use disorder: A psychodiagnostic perspective

Alcohol use disorder (AUD) is a chronic relapsing disorder with high early rehospitalization rates. Valid instruments assessing dependence severity, craving, and relapse risk are fundamental for follow-up. The aim of this study was to examine the psychometric properties of the Hungarian Severity of Alcohol Dependence Questionnaire (SADQ), Multidimensional Alcohol Craving Scale (MACS), Penn Alcohol Craving Scale (PACS) and Alcohol Relapse Risk Scale (ARRS), and test their utility in predicting relapse risk and rehospitalization. Ninety-three AUD inpatients were assessed on hospitalization day 8 with SADQ, AUDIT, MACS, PACS, visual analogue scale (VAS) of craving and ARRS, and followed for 3 months. Internal consistency and concurrent validity were evaluated. We tested whether SADQ and ARRS predicted 3-month rehospitalization (logistic regressions), whether MACS and PACS predicted relapse risk (linear regression). The 3-month rehospitalization rate was 21.5%. All scales showed high internal consistency (α = 0.86-0.94) and concurrent validity. After item reduction based on item-total correlations, the 27-item ARRS (ARRS-27) showed higher internal consistency, and predicted rehospitalization (OR = 1.12, 95% CI: 1.05-1.18; p < 0.001). Higher SADQ scores predicted rehospitalization (OR = 1.05, 95% CI: 1.02-1.09; p = 0.007). MACS (p < 0.001) and PACS (p = 0.046) scores predicted ARRS scores (R² = 0.53), with MACS showing the stronger effect. Based on our preliminary results, the Hungarian SADQ, MACS, PACS, and ARRS-27 are reliable instruments with predictive value for relapse risk and rehospitalization in AUD. Their combined use might support a treatment and relapse-prevention algorithm, improving clinical care and guiding future research.