Joint association of continuity of care and medication adherence with psychiatric readmission in mood disorders: a nested case–control study

Background

Psychiatric readmission imposes substantial clinical and economic burden in mood disorders. Medication adherence and continuity of outpatient care are both considered modifiable post-discharge factors but are typically examined in isolation. Whether their joint pattern differentially predicts readmission, and whether the pattern differs between bipolar disorder and depression, remains unclear.

Aims

To evaluate the joint association of adherence and continuity of care with psychiatric readmission in mood disorders, and to compare patterns between bipolar disorder and depression.

Methods

We conducted a nested case–control study using the Korean National Health Information Database (2009–2024), covering the entire population. Adults with an index hospitalisation for mood disorders (ICD-10 F30–F39) were followed from discharge. Cases were readmission episodes occurring >30 days after discharge; controls were matched 1:1 on age, sex, follow-up duration, mood disorder type, and number of prior admissions, using incidence density sampling. Adherence was measured by the proportion of days covered (PDC) for mood stabilisers (bipolar disorder) or antidepressants (depression). Continuity was measured by usual provider continuity (UPC), the share of outpatient visits to the most-visited provider. Each was categorised as low/medium/high (PDC <0.5/0.5–<0.8/≥0.8; UPC <0.5/0.5–<0.75/≥0.75), yielding nine combined exposure groups. Conditional logistic regression estimated adjusted odds ratios (ORs) with 95% confidence intervals (CIs), adjusting for income, residential area, and comorbidities. Subgroup analyses by mood disorder type were prespecified.

Results

The analytic sample comprised 102,252 episodes (51,126 cases, 51,126 controls; mean age 49.0 years, 54.9% male; 31.9% bipolar, 68.1% depression; median follow-up 6.4 months). Compared with the low-adherence/low-continuity reference, the combined high–high group had the lowest readmission risk (PDC ≥0.8 + UPC ≥0.75: OR 0.73, 95% CI 0.70–0.76). Among low-adherence patients, higher continuity remained protective (UPC ≥0.75: OR 0.80, 95% CI 0.77–0.82). Strikingly, high adherence combined with low continuity was associated with markedly higher risk (PDC ≥0.8 + UPC <0.5: OR 2.09, 95% CI 1.92–2.27).

Patterns diverged by diagnosis. In bipolar disorder, both high adherence and high continuity were independently protective (combined high–high: OR 0.56, 95% CI 0.52–0.60). In depression, higher continuity was consistently protective across adherence strata, whereas high adherence with low continuity was associated with substantially elevated risk (OR 2.81, 95% CI 2.53–3.11).

Conclusions.In this nationwide cohort, the joint pattern of medication adherence and outpatient continuity of care was associated with psychiatric readmission, with notable differences by diagnosis. In bipolar disorder, both higher adherence and higher continuity acted as independent protective factors, consistent with the relatively homogeneous clinical course. In depression, continuity was the dominant protective factor across adherence levels, while high adherence combined with low continuity was paradoxically associated with markedly elevated risk. This may reflect the clinical heterogeneity of depression and the possibility that frequent provider switching despite sustained medication use marks more severe or treatment-resistant illness, rather than direct harm from low continuity. Service models integrating adherence support with stable provider continuity may help reduce readmission, with diagnosis-specific tailoring.

References

No references