Background: Alcohol use disorder (AUD) is a relapsing-remitting disease that accounted for a sizable proportion of all-cause adult inpatient stays.
Objectives: To determine the predictors of any and multiple readmissions to inpatient care for AUD within 5 years of the index admission.
Methods: This retrospective, register-based cohort study assessed consecutive patients with AUD admitted to a publicly-funded inpatient service between January 2007 and December 2014. Binary logistic regression was used to determine independent predictors for readmissions based on relevant demographic, clinical, and treatment variables that showed significant differences (p < 0.05) on univariate analysis.
Results: Among 938 patients (age 35.9 ± 10.3 years; duration of alcohol use 159.6 ± 104.5 months; dual diagnosis 19%; comorbidity of substance use disorder 49.3%; medical disorder 34.8%, 299 (31.9%) and 115 (12.3%) had any and multiple readmissions, respectively. Comorbid "severe mental illness" (Odds ratio [OR] 1.99, 95% confidence interval [CI] 1.11-3.57) and urban residence (OR 1.58, 95% CI 1.13-2.18) increased the odds of any readmission; "Improved" status at discharge (OR 0.51, 95% CI 0.35-0.72) during index hospitalization reduced odds of readmission. Additionally, any medical or psychiatric comorbidities increased (OR 2.23, 95% CI 1.26-3.97), and comorbid substance use disorder decreased (OR 0.41, 95% CI 0.19-0.89) risk of multiple readmissions.
Conclusions: Clinicians could identify patients at-risk for any and multiple readmissions during the index hospitalization. A policy aimed at reducing the risk of rehospitalization, healthcare cost, and stigma should pay attention to the predictors of readmission. Such policy should further benefit resource-limited settings.
Keywords: Alcohol use disorder; inpatient care; predictors; readmission.