Key result
Neither the READMIT index (AUC 0.48) nor a new 36-predictor model (AUC 0.58) achieved acceptable discrimination (AUC ≥0.70) for 30-day depression readmission in 24,984 Danish patients.
Why the study?
The broad READMIT risk index may not be applicable to specific diagnoses like major depressive disorder, warranting validation and development of a specific 30-day readmission prediction model.
Does the READMIT risk index or a newly developed registry-based prediction model accurately predict 30-day psychiatric readmission in patients discharged after hospitalization for major depressive disorder?
Cohort (n=24,984)
Yes
Does the READMIT risk index or a newly developed registry-based prediction model accurately predict 30-day psychiatric readmission in patients discharged after hospitalization for major depressive disorder?
Neither the existing READMIT risk index nor a newly developed prediction model utilizing 36 registry-based variables achieved acceptable discriminatory ability (AUC ≥0.70) to predict 30-day readmission in patients discharged after treatment for major depressive disorder.
Answers are grounded in this paper and cite their sources.
Neither registry model discriminates 30-day readmission; leaves open need for richer predictors in post-discharge depression care.
When discharging patients, identifying those with a high risk of short-term readmission is crucial, as these readmissions may disrupt patient recovery and increase healthcare costs. The READMIT risk index has been developed to address this need across all psychiatric diagnoses. However, a broad model might not be applicable to specific diagnoses, e.g., major depressive disorder. We aimed to investigate the predictive validity of the READMIT risk index in a cohort of Danish patients suffering from major depression, and to develop an alternative prediction model for the risk of readmission within 30 days. Based on national register data, we included 24,984 patients discharged from a Danish psychiatric hospital after treatment for major depression. The proportions of patients who experienced readmission within 30 days were 8% (n=1969) with major depression, 13% (n=3343) with any psychiatric diagnosis, and 19% (n=4718) with all-cause readmission. For readmission related to major depression, the predictive performance of the READMIT index, defined as area under the receiver operating characteristic curve (AUC), was 0.48 (95% CI:0.47;0.49), whereas the prediction model developed in this study showed an AUC of 0.58 (95% CI:0.57;0.59). Neither the READMIT risk index nor a newly developed alternative prediction model based on individual-level variables, achieved an acceptable AUC, i.e., 0.70 or above, and were therefore not able to appropriately predict the risk of readmission within 30 days for patients with major depressive disorder. However, organizational factors and factors related to capacity in the Danish psychiatric system could potentially improve the results. • The READMIT index is not suitable for Danish major depression patients. • With 36 well-known predictors it was not possible to reliably predict readmission. • After treatment for major depression 8% of patients were readmitted within 30 days.
No takes yet. Share an insight, caveat, or question.
Grøntved et al. (2026) conducted a cohort in Major depressive disorder (n=24,984). READMIT risk index and newly developed 36-predictor clinical prediction model was evaluated on Acute hospital readmission related to major depressive disorder within 30 days of discharge (AUC 0.48 (READMIT index); CV AUC 0.58 (newly developed model), 95% CI 0.47-0.49). Neither the READMIT index (AUC 0.48) nor a new 36-predictor model (AUC 0.58) achieved acceptable discrimination (AUC ≥0.70) for 30-day depression readmission in 24,984 Danish patients.
Synapse has enriched 10 related papers. Consider them for comparative context: