Hospital Discharge Patterns and Outcomes Among Patients Hospitalized for Nonfatal Opioid Overdose in Arizona, 2016-2024

Introduction: Nonfatal opioid overdoses remain a significant public health issue. However, post-hospital discharge pathways and how they differ based on patient factors and co-occurring mental health conditions are still not well understood in Arizona. This dissertation focuses on three objectives: (1) identify predictors of discharge disposition after overdose hospitalization over time, (2) measure how discharge disposition influences hospital resource use (length of stay [LOS], total charges), and (3) explore the relationships between specific mental health categories and discharge outcomes among adults hospitalized for nonfatal opioid overdose from 2016–2024.Methods: A retrospective analysis was conducted using statewide discharge data from the Arizona Department of Health Services. Adults (≥18 years) with ICD-10 overdose codes T40.0X–T40.6X were included; emergency department discharges, ages 0-17, deaths, and law-enforcement discharges were excluded. The final number of observations included was 58,448. Discharge outcomes were routine, discontinued care, psychiatric hospital, health care facility, or home health. Multivariable logistic models estimated adjusted odds for each discharge type and mental illness. Multiple linear models evaluated log-transformed total charges, and LOS models assessed associations for stays >0 days, with all three studies adjusting for sociodemographic, clinical, and hospitalization factors. Analyses used SAS 9.4. Results: From 2016 to 2024, routine discharge was most common (51.5%), followed by health care facility (23.6%), home health (16.4%), discontinued care (4.6%), and psychiatric hospital (3.8%). The average length of stay (LOS) was 6.4 days (median 4.0); the average total charges were $87,913 (median $54,880). After adjusting for covariates, discharges due to discontinued care increased during periods 2 and 3, while discharges to psychiatric hospitals, health care facilities, and home health remained consistent over time. Compared with routine discharge, transfer to a health care facility, home health, or psychiatric hospital was linked to significantly higher charges and longer LOS, whereas discontinued care was associated with lower charges and shorter LOS. After adjusting for covariates, the variation of mental illness was shown across discharge outcomes. Discussion: Across three studies, differences were identified in hospital and clinical variables related to nonfatal opioid overdoses. These findings highlight opportunities to enhance discharge planning, incorporate behavioral health services, and improve care transitions for patients with co-occurring mental illness and SUDs in Arizona.