Transdiagnostic trajectories of acute response and readmission after electroconvulsive therapy: a retrospective cohort study in South Korea

Real-world evidence on electroconvulsive therapy (ECT) derives largely from Western cohorts with major depressive disorder (MDD), leaving diagnosis-specific long-term outcomes in refractory, heavily medicated patients less well characterized. We examined whether acute ECT response predicts readmission and whether response patterns diverge from postdischarge outcomes across diagnoses. In this retrospective cohort study at a tertiary referral center in South Korea (2023–2025), patients aged ≥ 16 years with MDD, bipolar affective disorder (BPAD), or schizophrenia who initiated an acute-phase ECT course were followed for transition to maintenance ECT (mECT) and psychiatric readmission. Outcomes included acute response and remission (Clinical Global Impression-Severity and diagnosis-specific scales), cognition (Montreal Cognitive Assessment), and readmission-free survival. Cox models included acute response and mECT as a time-dependent covariate. Among 135 patients (75 MDD, 25 BPAD, 35 schizophrenia), the acute response rate was 59.3% (95% CI, 50.8–67.2) and differed by diagnosis ( p = 0.001): BPAD 84%, MDD 61%, schizophrenia 37%, a pattern that persisted in sensitivity analyses. Within-group symptom reductions were large (Cohen d , 1.7–2.8), and cognition was stable at the group level. During a median follow-up of 175 days, 37 patients (27%) were readmitted. Acute responders had about half the hazard of readmission (adjusted hazard ratio [aHR], 0.49; 95% CI, 0.24–1.01). After adjustment for acute response, mECT, age, and sex, BPAD had a higher estimated readmission hazard than MDD (aHR, 2.11; 95% CI, 0.89–5.00), whereas schizophrenia had a similar hazard (aHR, 1.04; 95% CI, 0.39–2.77). Transition to mECT, least frequent in BPAD (28% vs. 54% in schizophrenia; p = 0.083), was associated with a nonsignificantly lower readmission hazard (aHR, 0.55; 95% CI, 0.25–1.22). Between-group differences in readmission-free survival were not significant (log-rank p = 0.248). Acute responders had a lower readmission hazard, and patients with BPAD had a higher estimated hazard than those with MDD after adjustment for acute response and mECT, although neither association was statistically significant. These findings suggest that acute improvement alone may not ensure durable stability in BPAD, warranting prospective evaluation of relapse prevention strategies in this group.