Routine follow-up care patterns after psychiatric hospital discharge and associations with rehospitalization and...

The risk of psychiatric readmission and death is high in the year after discharge. Existing posthospitalization interventions show modest and sometimes inconsistent average effects, suggesting that optimizing routine follow-up care may require identifying clinically meaningful heterogeneity in routine care practice. This study aims to identify latent subgroups of follow-up care patterns during the 12 months after psychiatric discharge and examine their associations with patient characteristics, rehospitalization, and all-cause mortality.

Using provincial linkable health administrative databases, we investigated a cohort of 3467 patients discharged after psychiatric hospitalization. Latent class analysis was used to identify subgroup memberships of follow-up care. Multivariable analyses estimated subgroups' associations with patient social and clinical correlates, and survival analyses estimated subgroups' associations with rehospitalization and all-cause mortality.

Five distinct subgroups of follow-up care patterns were identified, including low to high quality of care, diversified care and high access to primary care, with two subgroups having frequent emergency (ED) use. Serious mental disorders and prior hospitalizations were concentrated in the high follow-up care subgroups, where rehospitalization hazards were significantly elevated and approached one-third. Mortality did not differ across subgroups.

Follow-up care after psychiatric discharge is heterogeneous, reflecting patients' clinical risk profiles and outpatient contexts. Rehospitalization was driven primarily by clinical severity, while frequent ED use signalled unmet follow-up care needs, demonstrating two distinct post-discharge dynamics with different implications for quality monitoring and service improvement. These findings support subgroup-informed post-discharge optimization and may help explain the average modest effects of transitional interventions observed in prior research. Both carry direct relevance for policy and practice.