Binary logistic regression of current c-AWS.

Background Early recognition of the complicated form of alcohol withdrawal syndrome (c-AWS) is critical. The Prediction of Alcohol Withdrawal Severity Scale (PAWSS) was developed for the risk analysis of the development of c-AWS. According to the kindling mechanism, the history of previous c-AWS has a pivotal role in the development of current c-AWS. The aims of this study were to reveal (1) the psychometric characteristics of the PAWSS among patients with alcohol withdrawal syndrome (AWS) and alcohol dependence syndrome (ADS) and (2) the role of kindling mechanism in the development of c-AWS by using the PAWSS. Methods This study enrolled 70 inpatients with ADS and AWS. The severity of dependence was measured using the Alcohol Use Disorder Identification Test and the Severity of Alcohol Dependence Questionnaire. Statistical analyses were performed using receiver operating characteristic (ROC) analysis, binary logistic regressions, and for the inter-rater reliability analysis Cohen’s Kappa coefficient was calculated. Results ROC analysis showed that > 6 is the optimal cutoff point for the Hungarian version of the PAWSS. In the case of predictive validity, higher PAWSS score ( p < 0.001) predicted current c-AWS. Furthermore, the history of c-AWS ( p < 0.001) was a significant variable for current c-AWS. The Cohen’s Kappa coefficient resulted in being 1. Conclusions The probability of current c-AWS was 12 times higher among patients with PAWSS scores of 6 or higher. The chance of current c-AWS was almost 7 times higher in the case of history of c-AWS. These findings suggest that the Hungarian version of PAWSS is a valid and reliable clinical tool for assessing the risk of c-AWS, and highlight the importance of the kindling mechanism in the background of c-AWS.