ObjectiveTo assess the magnitude of suicide rates in the first week and first month postdischarge following psychiatric hospitalisation.DesignMeta-analysis of relevant English-language, peer-reviewed papers published in MEDLINE, PsycINFO or Embase between 01 January 1945 and 31 March 2017 and supplemented by hand searching and personal communication. A generalised linear effects model was fitted to the number of suicides, with a Poisson distribution, log link and log of person years as an offset. A random effects model was used to calculate the overall pooled rates and within subgroups in sensitivity analyses.Outcome measuresSuicides per 100 000 person years in the first week and the first month after discharge from psychiatric hospitalisation.ResultsThirty-four included papers comprised 29 studies that reported suicides in the first month postdischarge (3551 suicides during 222 546 patient years) and 24 studies that reported suicides in the first week postdischarge (1928 suicides during 60 880 patient years). The pooled estimate of the suicide rate in the first month postdischarge suicide was 2060 per 100 000 person years (95% CI=1300 to 3280, I2=90). The pooled estimate of the suicide rate in the first week postdischarge suicide was 2950 suicides per 100 000 person years (95% CI=1740 to 5000, I2=88). Eight studies that were included after personal communication had lower pooled rates of suicide than studies included after data extraction and there was evidence of publication bias towards papers reporting a higher rate of postdischarge suicide.ConclusionAcknowledging the presence of marked heterogeneity between studies and the likelihood of bias towards publication of studies reporting a higher postdischarge suicide rate, the first week and first month postdischarge following psychiatric hospitalisation are periods of extraordinary suicide risk. Short-term follow-up of discharged patients should be augmented with greater focus on safe transition from hospital to community care.PROSPERO registration numberPROSPERO registration CRD42016038169

analysis of suicide rates among people discharged from non-psychiatric settings after presentation with suicidal thoughts or behaviours.Objective: To quantify the suicide rate among people discharged from nonpsychiatric settings after presentations with suicidal thoughts or behaviours. Method: Meta-analysis of studies reporting suicide deaths among people with suicidal thoughts or behaviours after discharge from emergency departments or the medical or surgical wards of general hospitals. Results: A total of 115 studies reported 167 cohorts and 3747 suicide deaths among 248 005 patients during 1 263 727 person-years. The pooled suicide rate postdischarge was 483 suicide deaths per 100 000 person-years (95% confidence interval (CI) 445-520, prediction interval (PI) 200-770) with high between-sample heterogeneity (I 2 = 92). The suicide rate was highest in the first year postdischarge (851 per 100 000 person-years) but remained elevated in the long term. Suicide rates were elevated among samples of men (716 per 100 000 person-years) and older people (799 per 100 000 person-years) but were lower in samples of younger people (107 per 100 000 person-years) and among studies published between 2010 and 2018 (329 per 100 000 person-years). Conclusions: People with suicidal thoughts or behaviours who are discharged from non-psychiatric settings have highly elevated rates of suicide despite a clinically meaningful decline in these suicide rates in recent decades. Summations• The pooled suicide rate among patients with suicidal thoughts or behaviours following discharge from a non-psychiatric setting was 483 per 100 000 person-years (95% CI 445-520).• Suicide rates were lower among samples of younger people (<25 years of age) and among more recently published studies, but were higher in the first year after discharge, among older people (>55 years of age), and men.• All patients presenting to emergency departments and general hospitals with suicidal thoughts or behaviours warrant a careful assessment and individualized treatment planning. Considerations• High between-study heterogeneity limits the extent to which the results can be regarded as generalizable.• The meta-analysis could not examine a range of clinical factors that might contribute to variation in suicide rates.• Few studies from low-and middle-income countries could be included.

BackgroundPeople discharged from in‐patient psychiatric facilities have highly elevated rates of suicide, and there is increasing concern about natural mortality among the seriously mentally ill.MethodA meta‐analysis of English‐language, peer‐reviewed longitudinal studies of mortality among patients discharged from in‐patient psychiatric facilities was conducted using papers published in MEDLINE, PsycINFO or EMBASE (from 1 January 1960 to 1 April 2018) located using the terms ((suicid*).ti AND (hospital OR discharg* OR inpatient OR in‐patient OR admit*)).ab and ((mortality OR outcome* OR death*) AND (psych* OR mental*)).ti AND (admit* OR admis* OR hospital* OR inpatient* OR in‐patient* OR discharg*).ab. Pooled mortality rates for aggregated natural and unnatural causes, and the specific causes of suicide, accident, homicide, vascular, neoplastic, respiratory, gastrointestinal, infectious and metabolic death were calculated using a random‐effects meta‐analytic model. Between‐study heterogeneity was investigated using subgroup analysis and metaregression.ResultsThe pooled natural death rate of 1128 per 100 000 person‐years exceeded the pooled unnatural deaths of 479 per 100 000 person‐year among studies with varying periods of follow‐up. Natural deaths significantly exceeded unnatural deaths among studies with a mean follow‐up of longer than 2 years, and vascular deaths exceeded suicide deaths among studies with mean period of follow‐up of 5 years or longer.ConclusionSuicide may be the largest single cause of death in the short term after discharge from in‐patient psychiatric facilities but vascular disease is the major cause of mortality in the medium‐ and long‐term.

No abstract

No abstract