Predictors of readmission to substance-related disorder treatment over a five-year period

Treatment readmission is frequent among patients with substance-related disorders (SRDs). This study aimed to identify clinical, sociodemographic and service use predictors of readmission to SRD treatment within a 5-year period (2017-2022) following initial treatment episodes.

Data from 8277 patients showing at least one SRD treatment episode in specialized addiction treatment centers in 2013-2017 (Quebec, Canada) were merged with provincial health administrative databases (1996-2022). Readmission predictors were tested using Cox proportional hazards to measure patient characteristics, and SRD treatment over the previous 8 years (2009-2017). Logistic regression assessed other service use for the 12 previous months, as co-occurring disorders could influence readmission.

Over the 5 years following their last SRD treatment, 36 % of patients were readmitted. Were more likely to be readmitted: men, individuals who were younger, unemployed, not living in rural areas, had chronic SRDs, alcohol-related disorders, common mental disorders, 5+ prior SRD care episodes, got residential treatment, showed high dropout rates, and those who waited <30 days to access treatment. Readmission was likelier in patients who had access to psychiatrists and psychosocial services, had high continuity of physician care, and used more acute care - especially emergency departments.

Patients with more social and health issues and high prior and diversified service use showed higher risk of readmission. This suggests that care could be improved significantly to reduce readmission, especially through better detection and need assessment of patients with multiple prior SRD treatment episodes, dropouts and acute care use, and by increasing long-term follow-up care.