Suicide risk after discharge from in-patient psychiatric care: A 15-year retrospective cohort study of individual patient data

Suicide rates are known to be increased in patients after discharge from in-patient psychiatric treatment. However, evidence on risk factors for suicide within this patient group are contradictory. Thus, this study aims to investigate suicide after discharge from a sizeable psychiatric care facility to determine associated risk factors. Data on individual patient level from a 15-year single-centre cohort were linked to data from the national death registry and cumulative incidence rates were calculated applying competing risk models. Independent variables included the patients' sex, age at admission, diagnosis, and length of admission. For each of these factors, subdistribution hazards ratios were calculated using a Fine-Gray model. In our sample of 18,425 discharges, when using patients with the diagnosis of substance-use-disorders as a comparator, a significant increase in hazard of post-discharge suicide for male sex (SHR = 1.67;p = 0.037) as well as the discharge diagnoses of affective disorders (SHR = 3.56;p = 0.017) and neurotic stress and somatoform disorders (SHR = 3.73;p = 0.024) were found. Interestingly, the hazard of suicide significantly decreased in more recent discharges (SHR = 0.93;p = 0.006). No statistically significant association of the length of admission with the suicide risk was found (SHR = 0.98;p = 0.834). Limitations. Suicides may have been mis-identified as natural death in the national death register. Male sex and distinct diagnoses were associated with an increased risk for suicide after discharge from a psychiatric care institution. The markedly increased suicide risk within this patient collective highlights the need for the development of tools to assess suicidal behaviour in this group of patients reliably.