Factors associated with prolonged psychiatric intensive care unit admission at Somdet Chaophraya Institute of Psychiatry, Thailand

Aim: Our objective is to determine factors associated with prolonged psychiatric intensive care unit (PICU) admission at Somdet Chaophraya Institute of Psychiatry (SCIP).<br/> Method: We reviewed the medical records of patients admitted to the PICU between 1 July 2016 and 31 December 2016 meeting PICU admission criteria. We defined prolonged admission as PICU clinical stay of longer than five days. We then performed univariate binary logistic regression analysis to analyse the association between factors and prolonged PICU admission. Factors with p-value <0.25 were then included in backward stepwise multiple logistic regression analysis.<br/> Results: We included 293 cases then excluded 42 of them using our exclusion criteria (30 were discharged by other means and 12 cases were due to inability to obtain essential information). Of 251 studied cases, 112 cases (44.6%) had prolonged PICU admission. Univariate logistic regression analysis found that a history of prior PICU admission, compulsory admission, treatment with electroconvulsive therapy (ECT), and primary diagnosis were associated with prolonged PICU admission with p-value <0.25. Using multiple logistic regression analysis, factors associated with prolonged PICU admission were: compulsory admission (adjusted odds ratio 2.45, 95% CI 1.06–5.69 when adjusted with treatment with ECT) and treatment with ECT during PICU admission (adjusted odds ratio 9.01, 95% CI 2.57–31.59 when adjusted with compulsory admission).<br/> Conclusions: Factors associated with prolonged PICU admission at SCIP were compulsory admission and treatment with ECT during PICU admission. We propose that further study of these two groups should provide clues on how to improve treatment in PICU.