Variations in involuntary admission rates at three psychiatry centres in the Dublin Involuntary Admission Study (DIAS): Can the differences be explained?

Abstract

Involuntary psychiatric admission is an established practice for patients who are acutely or severely mentally ill but the factors contributing to involuntary (as opposed to voluntary) admission are not fully clear. Nor is it clear why rates of involuntary admission often vary between hospitals within the same jurisdiction. We studied all admissions, voluntary and involuntary, in three inpatient psychiatry units in Dublin, Ireland, which cover a population of 552,019 people, over a one-year period (1 July 2014 until 30 June 2015, inclusive), as part of the Dublin Involuntary Admission Study (DIAS). During the study period, there was a total of 1136 admissions to these three units, of which 17% were involuntary for all or part of their admission. The overall admission rate (205.8 admissions per 100,000 population per year) was lower than the national rate (387.9) but this varied substantially across the three units studied. On multi-variable analysis, involuntary admission status was associated with male gender, being unmarried, and a diagnosis of schizophrenia, and was not significantly associated with age, occupation or which inpatient unit the person was admitted to. We conclude that variations in involuntary admission rates between different psychiatry admission units in Dublin are significantly explained by patient-level variables (such as gender, marital status and diagnosis) rather than centre-level variables, but that much of the variation in admission status between patients remains unexplained. Future, multi-level research could usefully focus on other patient-level factors of possible relevance (e.g. symptom severity), centre-level factors (e.g. local mental health service resourcing) and community-level factors (e.g. socio-economic circumstances in different areas) in order to further elucidate unexplained variance in admission status between patients.