708. COMPARISON OF FEATURES OF RELAPSE RISK BETWEEN ALCOHOL-DEPENDENT INDIVIDUALS IN JAPAN AND AUSTRALIA THROUGH USE OF VALIDATED SELF-RATING SCALE: THE ALCOHOL RELAPSE RISK SCALE

Abstract Background Alcohol use disorder (AUD) is a prevalent, chronic condition with potential serious consequences. The Alcohol Relapse Risk Scale (ARRS) is a multi-dimensional, 32-item self-reported scale developed in Japan which has shown correlation in the clinical setting to predict alcohol relapse. The scale is divided into five subscales: stimulus-induced vulnerability, emotionality problems, compulsivity for alcohol, negative expectancy for alcohol, positive expectancy for alcohol, and insight into mental conditions. Aims & Objectives The present study aimed to study the applicability of ARRS in the Australian setting for patients who fit the criteria of AUD and are engaged with local drug health services through the inpatient detoxification unit or general hospital consultation-liaison service. Method A total of 101 patients who fulfilled the DSM-5-TR diagnostic criteria for AUD were recruited in South West Sydney Local Health District. Participants were screened using the ARRS score (English version) and followed up after 4 weeks to assess for alcohol relapse within the period. Internal consistency was analysed using the Cronbach’s a coefficients, and predictive validity was measured using Pearson’s correlation coefficient and logistic regression. Furthermore, ROC analysis was conducted to find potential cutoff points of ARRS scores which could divide patients into varying risk groups. Results The use of ARRS in the Australian setting resulted in reasonable internal consistency (Cronbach’s a = 0.86 for the total score) and trended towards positive correlation of score with relapse risk, although statistical significance of predictive validity was only achieved in the IM subscale. ROC analysis yielded a suggested cutoff score of 71.5 (maximum score = 96), which resulted in a sensitivity of 0.8 and specificity of 0.42 when applied to the Australian setting. Discussion & Conclusions There was a general similarity between the Australian and Japanese data sets, including their internal consistency and predictive validity, except the engagement setting, where the Japanese study population had significant recruitment in the outpatient setting. In conclusion, ARRS is a promising tool for use in the Australian setting to aid with relapse risk prediction.