The suicide risk of discharged psychiatric patients.

BACKGROUND: The suicide risk of psychiatric patients fluctuated along the course of their illness and was found to be high in the immediate post-discharge period in some settings. The epidemiology and psychiatric services for the suicide population in Hong Kong have differed from those of the West (i.e., low youth suicide rate, high elderly suicide rate, high female/male ratio, and heavily government-subsidized psychiatric service). This study examined the suicide rates within a year of discharge from psychiatric inpatient care in Hong Kong.

METHOD: Discharges from all psychiatric hospitals or psychiatric wards in general hospitals in Hong Kong from 1997 through 1999 were followed up for suicides (ICD-9, E950-E959) and "undetermined" causes of deaths (E980-E989) by record linkage with the Coroner's Court until their deaths or Dec. 31, 2000. The suicide rates (/1000 person-years at risk) and standardized mortality ratios (SMRs; assigning a value of 1 to the same age- and sex-specific suicide rates in the general population) were calculated.

RESULTS: 21,921 patients (aged over 15 years) were discharged from psychiatric hospitals from 1997 through 1999. Two hundred eighty patients committed suicide within 1 year of discharge; 85 suicides (30%) occurred within 28 days after discharge. The SMRs for suicide in the first 28 days after discharge were 178 (95% CI = 132 to 235) for females and 113 (95% CI = 86 to 147) for males. These rates were 4.0 (95% CI = 2.7 to 5.6) times higher for females and 4.6 (95% CI = 3.2 to 6.3) times higher for males than the rate in the rest of the year. Young adults had higher SMRs than the elderly. No specific diagnoses had higher suicidal risk than others. Calculations including undetermined causes of deaths (N = 53) gave similar results.

CONCLUSION: The immediate post-discharge period carries a high risk of suicide for psychiatric patients. The high-risk groups are young adults and females. No diagnosis appears to carry a particularly high risk.