Alcohol craving predicts relapse after residential addiction treatment

Aims: Replicate the previously reported association of elevated alcohol craving, measured by Penn Alcohol Craving Scale (PACS) during residential treatment, with post-treatment relapse and explore whether elevated craving scores 3 months post-treatment are also associated with subsequent relapse. Methods: Alcohol craving was assessed with the PACS on admission and at several time points post-treatment in 190 subjects with DSM-IV diagnosis of alcohol dependence admitted to residential treatment. Data about relapse to any drinking (primary outcome measure) was collected at 3, 6, 9 and 12 months after treatment. Cox regression models were used to determine whether PACS scores were associated with relapse. Statistical models were adjusted for meaningful demographic and clinical covariates. Results: Follow-up data was available for 149/190 (78%) of subjects. Elevated PACS scores at discharge were associated with increased relapse risk within the first 3 and 12 months after discharge (P = 0.032 and P = 0.045, respectively). Elevated PACS scores at 3 months were associated with increased risk of subsequent relapse within 12 months after treatment in contacted subjects (P = 0.034) and in the intent-to-treat analysis (P = 0.0001). Conclusions: Our findings indicate strong association of post-treatment relapse with elevated alcohol craving measured at treatment completion and at 3 months after treatment and justify the use of this measure to guide relapse-prevention efforts.