(80) Comparative Outcomes of Opioid Withdrawal Management for Patients Admitted with Bacteremia

Background & Introduction: Hospitalized patients with opioid use disorder (OUD) present challenges in patient care due to patient-directed discharge (PDD), incomplete treatment, and frequent readmission. Injection-related bacteremia is common and often requires long-term intravenous antibiotics. This study compares clinical outcomes for OUD patients admitted for bacteremia between those initiated on medications for OUD (MOUD) during hospitalization and those who received high dose opioid agonist therapy only (HDOAT).

Methods: This is a single-centered, IRB-approved, retrospective chart review study conducted at Temple University Hospital in Philadelphia, PA. Hospitalized patients with OUD and bacteremia admitted from Jan 1st, 2020, through Dec 31st, 2022, were included. Endocarditis was specifically excluded. Outcomes of interest included patient directed discharge (PDD) rates, 7 day (7d) and 30 day (30d) readmission rates and inpatient mortality. After manual review, 303 new cases from 383 encounters and 268 unique patients were identified. Odd ratios were calculated comparing encounters in which patients were initiated on MOUD and patients treated with only HDOAT. The MOUD group received either buprenorphine or methadone while the HDOAT group received no MOUD and were treated with oxycodone and hydromorphone for withdrawal management. Timing of MOUD initiation and HDOAT dose strengths were compared in subgroup analyses.

Results: Of the 383 encounters,149 declined MOUD during hospitalization and were managed only with HDOAT, while 234 accepted MOUD prior to discharge. Comparing the HDOAT and MOUD groups, patients who initiated MOUD experienced longer lengths of stay (LOS) (3d [0,63] v 8d [1,111]), lower PDD rates (61% v 38%; OR 2.6, p< 0.0001), and similar 7d and 30d readmission rates (20% v 16%; OR 1.4, p=0.2; and 38% v 33%; OR 1.3, p=0.3). There were no significant differences in inpatient mortality. Additionally, we compared encounters in which MOUD was initiated within 72hrs (n=168) and after 72hrs (n=66) of hospitalization. There were no significant differences in PDD rates (p=0.2), 7d readmission (p=0.5), 30d readmission (p=0.6), or inpatient mortality (p=0.6). Furthermore, we stratified HDOAT patients based on dose strength received, converted to morphine milligram equivalents (MMEs), into three tiers: low ( < 240 MMEs/day; n=59), medium (240-480 MMEs/day; n=43) and high (>480 MMEs/day; n=47) dose groups. No statistical differences were found in rates of PDD, 7d readmission, 30d readmission or inpatient mortality between the dose tiers. Finally, looking at the 303 cases, the rate of antibiotic completion was higher in the MOUD group (35.3% v 23.5%; OR 1.78, p< 0.031), with some patients completing treatment in non-contiguous encounters.

Conclusion & Discussion: MOUD demonstrated better overall outcomes compared to HDOAT with lower rates of PDD and higher antibiotic completion rates, regardless of initiating within or after 72 hours of admission. We did not find additional benefits to early vs late initiation of MOUD in patients with bacteremia. This suggests providers have time to counsel and address patients’ concerns regarding MOUD. In patients who continue to decline MOUD despite counseling, HDOAT may be an alternate strategy as 23% of HDOAT patients still completed their antibiotic regimens, though 39% of these patients completed treatment in non-contiguous admissions. Furthermore, there were no differences in readmission rates between the MOUD and HDOAT groups. This may indicate that by providing HDOAT to manage withdrawal, patients were willing to return to continue treatment despite previous PDD. We conclude that MOUD should be prioritized and initiated in the hospital for patients with bacteremia. However, HDOAT may be an appropriate option in select patients who continue to decline MOUD. Further research into the characteristics of patients who decline vs accept MOUD should be investigated to target resources to encourage transition to MOUD. Additionally, comparison of outcomes for patients initiated on methadone vs buprenorphine should be investigated in this patient population.

References: Compton P, Aronowitz SV, Klusaritz H, Anderson E. Acute pain and self-directed discharge among hospitalized patients with opioid-related diagnoses: a cohort study. Harm Reduct J. 2021 Dec 16;18(1):131. doi: 10.1186/s12954-021-00581-6. PMID: 34915913; PMCID: PMC8679978.

Crepet A, Sargent C. Infections in people who inject drugs on the acute medical take. Clin Med (Lond). 2022 Sep;22(5):383-386. doi: 10.7861/clinmed.2022-0364. PMID: 36507822; PMCID: PMC9595000.

Herscher M, Fine M, Navalurkar R, Hirt L, Wang L. Diagnosis and Management of Opioid Use Disorder in Hospitalized Patients. Med Clin North Am. 2020 Jul;104(4):695-708. doi: 10.1016/j.mcna.2020.03.003. Epub 2020 May 12. PMID: 32505261.

Reed MK, Sarpoulaki N, Murali V, Weinstein LC, Zavodnick JH, Bowles J, Rising KL. A Qualitative Analysis of Hospitalist Perceptions of Self-Directed Discharge Among Inpatients with Opioid Use Disorder. Subst Use Misuse. 2023;58(7):881-888. doi: 10.1080/10826084.2023.2191698. Epub 2023 Mar 29. PMID: 36987991.