Predictors of readmission and emergency-department reattendance after hospitalisation for suicidal ideation or attempt: a six-year cohort study

Background and aims. Psychiatric admission for high-risk suicidal behaviour — suicidal ideation (SI) or suicide attempt (SA) — is frequent standard of care. Admission decisions rest largely on clinical judgement and on the imperative to mitigate emergency department post-discharge harm, which likely contributes to the high rate of inpatient referral. Yet the impact of hospitalisation on subsequent suicidal behaviour remains poorly understood, and suicide risk is exceptionally elevated in the weeks after discharge [1]. Randomised trials comparing hospitalisation, with discharge or intensive outpatient care, have shown no benefit on symptom burden or suicide risk [2]. By contrast, admission after a suicide attempt may transiently reduce the risk of a further attempt — an effect not seen with ideation alone [3]. This heterogeneity sustains uncertainty around the routine use of inpatient care. We set out to identify predictors of post-discharge utilisation in adults admitted for SI or SA, and to test whether the SI/SA distinction itself stratifies risk.

Methods. We conducted a retrospective cohort study of all consecutive admissions to a psychiatric inpatient unit in Lisbon between January 2019 and December 2024. From 1458 admissions, 408 motivated by SI or SA formed the analytical sample. Sociodemographic, clinical, treatment and post-discharge data were extracted from clinical records. Continuous variables were compared with Mann–Whitney U and categorical variables with chi-squared or Fisher’s exact tests. Three multivariable logistic regression models (for 90-day readmission, and for 30- and 90-day emergency-department [ED] reattendance) were fitted with identical covariates: age, sex, admission motive, substance use, prior admissions, length of stay, and depressive or personality disorder. Estimates are reported as odds ratios (OR) with 95% confidence intervals (CI); model fit by McFadden’s pseudo-R². Significance set at <0.05.

Results. Of 408 admissions (57.1% women; median age 42 years, IQR 27–55), 234 (57.4%) were for SI and 174 (42.6%) for SA. Within three months of discharge, 9.9% of patients had been readmitted, 29.1% had returned to the emergency department, and 20.9% had done so within the first month. The SI and SA groups proved indistinguishable on age, sex, length of stay and on every post-discharge outcome examined. Across the three multivariable models, the number of prior admissions was the only consistent independent predictor of post-discharge utilisation (OR≈1.25 per admission, 95%CI 1.10–1.43, p<0.001 for all three outcomes). Female sex independently predicted 90-day readmission (OR for male sex 0.40, 95%CI 0.17–0.94, p=0.036), and a shorter index stay predicted 90-day ED reattendance (OR 0.98 per day, 95%CI 0.96–1.00, p=0.024). Age, admission motive, substance use and primary diagnosis showed no association with any outcome.

Conclusions. In this six-year cohort, each prior admission raised the odds of subsequent readmission or ED reattendance by approximately 25%. Women and patients with shorter index stays were particularly vulnerable; the binary SI/SA distinction added no prognostic value. These findings support risk-tiered transitional care anchored on hospitalisation history, prioritising women and frequent admitters for intensified post-discharge support.

References

1. Chung DT, Ryan CJ, Hadzi-Pavlovic D, et al. Suicide rates after discharge from psychiatric facilities: a systematic review and meta-analysis. JAMA Psychiatry 2017;74(7):694–702.
2. Olfson M, Wall M, Wang S, et al. Short-term suicide risk after psychiatric hospital discharge. JAMA Psychiatry 2016;73(11):1119–1126.
3. Forte A, Buscajoni A, Fiorillo A, et al. Suicidal risk following hospital discharge: a review. Harvard Review of Psychiatry 2019;27(4):209–216.