Predictors for Extending Hospitalization Stay in Electroconvulsive Therapy Recipients With Bipolar Disorder, Manic Episodes

Abstract

We aim to discern the demographic predictors that may extend the hospitalization length of stay (LOS) for patients with bipolar disorder (BD), manic episodes managed with electroconvulsive therapy (ECT), and to study the impact of insurance and hospital characteristics on LOS. Methods We used the Nationwide Inpatient Sample (NIS, 2012-2014) from the United States hospitals and included 2,785 adult inpatients (mean age 51.3 ± 16.2 years) with a primary diagnosis of BD, manic episode, and managed with ECT. The median LOS of the sample population is 16 days, and the study inpatients were divided into subgroups: ≤16 days versus >16 days. The logistic regression model was used to find the odds ratio (OR) for the associations of demographic and hospital variables with inpatient stay >16 days versus ≤16 days. Results BD inpatients managed with ECT during their hospitalization had a mean LOS of 21.6 ± 22.1 days. About 48.65% (N = 1355) had LOS >16 days. Older adults (age >50 years) have 2.4 times higher odds (95% CI 2.06- 2.87) for hospital LOS >16 days compared to younger adults. Although a higher proportion of females received ECT (71.8%), males had two times higher odds (95% CI 1.59-2.27) for hospital LOS >16 days. BD inpatients covered by private insurance/self-pay were at 1.5 times higher odds (95% CI 1.27-1.77) for hospital LOS >16 days. In terms of hospital setting, ownership type and teaching status are significant predictors with inpatients managed in public and teaching hospitals at higher odds for LOS >16 days. Conclusions Older men and inpatients covered by private insurance/self-pay have a higher likelihood of extended hospitalization stay during ECT management of BD, manic episodes. The LOS is also influenced by hospital setting with patients managed in public teaching hospitals at higher odds of longer LOS compared to their counterparts.