Short-Acting Opioid Dose and Patient-Directed Discharge in Hospitalized Patients With Opioid Use Disorder

Importance: Patient-directed discharge, common among hospitalized patients with opioid use disorder (OUD) experiencing opioid withdrawal, is associated with poor outcomes. Increasingly, hospital-based clinicians are using short-acting opioids alongside methadone and buprenorphine to treat opioid withdrawal. However, no published studies have examined the association between this practice and patient-directed discharge. Objective: To examine associations between dosing of short-acting opioids and time to first dose of short-acting opioid with early patient-directed discharge (within 72 hours). Design, Setting, and Participants: This retrospective cohort study included hospitalizations involving adults with OUD who were treated with short-acting opioids between February 7 and December 31, 2024, using an opioid withdrawal order set in the electronic health record at an urban safety-net hospital. Statistical analysis was conducted between October 2024 and October 2025. Exposures: Total daily dose of short-acting opioids (oxycodone and hydromorphone) expressed in morphine milligram equivalents, and time from emergency department intake to first dose of short-acting opioid. Main Outcomes and Measures: The primary outcome was early patient-directed discharge, adjusted for age, race and ethnicity, gender, other substance use disorder, receipt of medications for OUD, and addiction care team consultation. Results: A total of 669 hospitalizations, representing 520 unique patients, were included. Patients were a mean (SD) age of 44.1 (12.2) years, with most identifying as male (531 of 669 [79.4%], with 122 of 669 female [18.2%] and 16 of 669 gender minority [2.4%]) and unhoused (555 of 669 [83.0%]). A total of 92 hospitalizations (13.8%) resulted in early patient-directed discharge. In the final adjusted model, each doubling of the patient's daily dose of short-acting opioids was associated with an 8% reduction in the hazard of early patient-directed discharge (adjusted hazard ratio, 0.92 [95% CI, 0.86-0.99]; P = .03). There was no statistically significant association between time to first dose of short-acting opioid and patient-directed discharge within 72 hours. Conclusions and Relevance: In this cohort study of hospitalized adults with OUD, treatment of opioid withdrawal with short-acting opioids exhibited a dose-dependent association with reduced hazard of early patient-directed discharge. These findings support the use of short-acting opioids alongside methadone and buprenorphine for inpatient treatment of opioid withdrawal.