Quantifying institutional-level length of stay variation among hospitalizations for schizophrenia in Ontario between...

Variations in the length and intensity of care delivery from person to person are to be expected, however, such variation can also result from institutional-level factors that may not be directly related to a patient's needs. The objective of this study is to provide a descriptive analysis quantifying the variation in the length of stay (LOS) that is attributable to institutional- and patient-level factors for Ontarians hospitalized with schizophrenia between 2014 and 2021. A retrospective cohort study was conducted using Ontario medical records from >100,000 adult inpatients who had been admitted to an Ontario hospital with a primary diagnosis of schizophrenia between fiscal years 2014 and 2021. The proportion of variation in inpatient LOS that was attributable to institutional-level factors was assessed using log-linear mixed-effects models. Large community, teaching, and specialty mental health hospitals were each modelled separately. These results are presented alongside descriptive analyses for additional context. Average LOS in large community hospitals (mean = 24.2 days, SD = 62.51 days) was lower than specialty mental health hospitals (mean = 85.4 days, SD = 262.9 days) and teaching hospitals (mean = 42.8 days, SD = 137.4 days). The highest proportion of institutional-level variation was seen in large community hospitals (29.3%), with specialty mental health hospitals (19.3%) and teaching hospitals (19.7%) reporting similar proportions of institutional-level variation. This analysis has identified differences in the proportion of inpatient LOS variation attributable to institutional-level factors between types of hospital. A larger percentage of institutional-level variation was seen in large community hospitals however, all hospital types exhibited institutional-level variation. These differences likely result from a combination of governmental and hospital-level factors, which may present an opportunity for policy and clinical interventions. Future research can assess the potential for evidence-based interventions such as bundled care pathways and supportive housing programs in addressing these factors.