P-868. Creation of a Novel OUD-Specific Antibiogram

Abstract Background People who inject drugs are host to a clinically distinct microbial ecology given the population’s high pathogen burden and repeated, incomplete antibiotic exposures. Syndromic antibiograms have a potential utility in this population, and have not to our knowledge been used in patients with opioid use disorder (OUD) and may inform clinically relevant practice changes.OUD Antibiogram TableThe completed antibiogram table comparing the OUD population we identified with the general institutional population.Hospitalizations for Substance Use-Associated Infections in Philadelphia, 2010-2021Citywide hospitalization trends for skin and soft tissue infection and bacteremia in patients with substance use disorder show clear upward trajectories. Methods We conducted a retrospective study of adults (≥18 years) admitted to Thomas Jefferson University Hospital or Jefferson Methodist Hospital during 2022-2023. Persons with OUD were identified by presence of OUD ICD diagnoses (F11.x ICD10 codes) and a positive fentanyl urine test, and positive cultures were included. Patients under 18 were excluded. Patient records were identified through EPIC and cross-referenced with Clinical Microbiology Laboratory data. We analyzed susceptibility patterns for common injection-related organisms (if ≥30 isolates present) and compared them with institutional antibiogram data after removing the OUD population.Detections in Overdose Decedents' Toxicology TestsThe prevalence of benzodiazepines versus xylazine in post-mortem toxicology in Philadelphia, showing xylazine's growing market share. Results 1807 patients met OUD criteria over the two-year period. Susceptibility data was obtained for Staphylococcus aureus, Streptococcus pyogenes, Pseudomonas aeruginosa, Proteus mirabilis, E. coli for both the OUD group and total group, and an antibiogram was created as per institutional protocol. Cochran–Mantel–Haenszel testing showed a significant dependence (p=2.35e-8) between OUD and susceptibility when controlling for antibiotic-organism pairs. TMX-SMX susceptibility amongst MSSA isolates was 97% (non-OUD) versus 92% (OUD)(p=.012), and amongst MRSA isolates was 85% (non-OUD) versus 80% (OUD)(p=.087). Significantly increased resistance patterns were seen amongst E. Coli isolates in the non-OUD cohort versus OUD (ampicillin-sulbactam 57% vs 43%, p=.006; TMP-SMX 72% vs 62%, p=.012, ampicillin 48% vs 3%, p=.004). 32% of all S. pyogenes isolates for the given culture sites were from persons with OUD. Conclusion A novel customized antibiogram that showed statistically and clinically significant differences in antibiotic susceptibility rates in the inpatient population with chart-identified OUD and identified valuable population level infection prevalence data. Disclosures All Authors: No reported disclosures