Type of study: Retrospective cohort study

The finding: Using four linked existing datasets, a cohort of patients was identified who had been admitted for psychiatric hospitalization between a period of 5 years in New South Wales, Australia and examined in the period of 24 months following their discharge from that index admission for subsequent ED presentations and readmissions to a psychiatric facility. On the primary outcome measure, it was found that 37% of these individuals had at least one ED presentation in the 24 months after this index admission, and of those, 28% in the first month, 50% in the first 2-5 months, and 80% in the 6-24 month period. Of the factors that increased likelihood of presenting to the ED following index admission, female gender (in the 2-5 mo and 6-24 mo interval), younger age, living in outer regional areas and socioeconomically disadvantaged areas, those with disorders of adult personality and behaviour, those with shorter lengths of stay, those with hospitalizations for non-psychiatric comorbidities, those with intellectual disability and those with drug and alcohol comorbidities. Of those discharged from psychiatry, 40% had at least one psychiatric readmission in the first 24 months.

Strength and weaknesses: The study design did not allow the cohort to be correctly identified, as it selected individuals who had been free of psychiatric admissions for at minimum two years prior to the beginning of the study period rather than the individuals who had their first lifetime psychiatric admission, as was its stated intention. The study also lacked clinical information about the individual’s visits that may also be contributing factors to acute psychiatric presentations. This aside, the study makes use of available data sets as a starting point to guide further research and identify additional service needs for subsets of the population.

Relevance: As demand for acute psychiatric services increases and the problem of ED boarding continuing to grow, it is increasingly important to identify factors both on the regional and national level that are driving a return to the ED or inpatient setting following an acute inpatient discharge, and to make use of existing platforms (such as the EMR) as an opportunity to capture this data. As suggested by the study, identifying sociodemographic factors that are strong predictors of ED presentations (such as intellectual disability, personality disorder, or drug and alcohol disorder) provide an opportunity to develop specialized resources in the community.

Type of study: Retrospective cohort study

The finding: Under pressure to reduce boarding of psychiatric patients, EDs often re-evaluate patients and may then discharge patients who were previously destined for inpatient psychiatry. The authors evaluated 3 cohorts of patients from a single ED: those admitted to psychiatry, those discharged from the ED, and those scheduled for admission who were subsequently discharged from the ED. Patients discharged after initially being scheduled for admission were more likely to re-present to the ED within 30 days (29% v 11-19%) and 12 months (54% v 31-41%) but there was no statistical difference in suicide attempt (5-7%) or all-cause mortality (2-3%). Overall, the practice of re-evaluation and discharge has grown increasingly common since 2015.

Strength and weaknesses: This study identifies practical, clinically relevant outcomes for ED patients. Their use of bed request data is a clever way to add nuance to the discussion of boarding patients and practices that have risen in response to increasing psychiatric demand in EDs. The author’s linkage to national mortality data minimizes the risk that outcomes were missed due to the study’s restriction to one institution. The authors do not well describe what the re-assessment process looks like at their institution. Nor did the authors adjust for confounding variables, even as the cohorts unsurprisingly varied in key respects including chief complaint, diagnosis, and degree of family support.

Relevance: Psychiatric patients often have longer lengths of stay and resource utilization while in the ED. Discharging patients early is not an adequate solution to that problem. Other care models are available to enhance the care of psychiatric boarders including ED-based integrated care.