Hospital characteristics associated with post-discharge suicide of severely depressed patients

PURPOSE: To explore risk factors, such as characteristics of psychiatrist, patient and hospital at index hospitalization, associated with depressed patients who committed suicide within 3 months of discharge using a case-control design. METHODS: By linking the Taiwanese nationwide mortality database and the National Health Insurance dataset, all hospitalized patients with major depression who committed suicide within a 90-day period post-discharge during the years 2002-2004 were selected as a study cohort (n=85). We randomly selected 425 cases (five for every case in the study cohort) that were matched with the study cohort in terms of age, gender and date of discharge as a control cohort. Cox proportional hazard regression was carried out to compute the adjusted 90-day survival rate after adjusting for other factors. RESULTS: The majority of suicide occurred 30 days after discharge from hospitals, with a mean of 29.9 days. The adjusted hazard for committing suicide after hospital discharge for patients who left on their own initiative was 2.85 times (95% CI=1.387-5.856, p=0.004) greater than for those who were discharged with doctors' approval. Furthermore, the adjusted hazard for patients who were discharged from medical centers was higher than for patients discharged from regional hospitals, by a multiple of 3.38 (95% CI=1.421-8.055, p=0.006). LIMITATIONS: Some patient-specific predictors of post-discharge suicide, such as suicidal ideation prior to admission, recent life events and social relationships with close relatives were not available. CONCLUSION: We conclude that, in addition to patient characteristics, hospital characteristics also constitute important risk factors for suicide within 90 days of discharge for depressed patients.