1643. Infectious Sequelae of Injection Drug Use at a County Safety-Net Hospital Decline Following Establishment of a Syringe Service Program

Abstract

BackgroundInfectious sequelae of injection drug use (ISIDU) and overdose are frequent but preventable among people who inject drugs (PWID). Syringe service programs (SSP) are an evidence-based harm reduction strategy to reduce incidence of ISIDU among PWID. Additionally, SSPs are noted to produce significant cost-savings for healthcare systems. Under current state legislation, Miami houses the only SSP in Florida, the IDEA SSP. This study builds on previous work characterizing morbidity and cost of ISIDU. This study sought to evaluate differences in admission rates and associated ISIDU costs at Jackson Memorial Hospital (JMH) in Miami before and after implementation of the IDEA SSP.MethodsRetrospective data collected from a chart review of patients hospitalized for ISIDU and overdose was used to evaluate morbidity and cost of ISIDU at JMH from December 1, 2015 to December 1, 2017, stratified by December 1, 2016—the opening of the IDEA SSP—as an index date. An algorithm utilizing ICD-10 codes for drug use and sequelae was used to identify PWID population. Specific infections investigated were: endocarditis, osteomyelitis, bacteremia- and/or -sepsis (BOS), and skin-and-soft-tissue-infections (SSTIs). Pearson’s chi-square test for independence used to report P-values for associations between infections and total charges using a 2-tailed t-test.Results726 admissions were identified during the study period, 328 PWID in the pre-index cohort vs. 398 in the post-index cohort. The median age of total sample was 45.24. 95.12% of the pre-index cohort were uninsured or had publicly-funded insurance vs. 96.48% post-index. Most ISIDU did not change significantly between pre-post cohorts, although bacteremia and sepsis declined significantly among opioid injectors (P = 0.026). Overdoses decreased significantly among PWID generally (57% decline pre-post; P = 0.0006), as well as for patients who inject opioids specifically (70% decline pre-post; P = 0.0034). Median cost declined by 20.5% among PWID, and 29.1% among opioid injectors in particular.ConclusionISIDU continues to represent significant morbidity for PWID in Miami-Dade County and substantial cost to the health system. Severe infections, including bacteremia and sepsis, declined significantly among opioid injectors, the PWID subset most strongly associated with local SSP services. This change following the establishment of a local SSP suggests direct effects on the frequency of hospital admissions for ISIDU. Despite local increases in drug use, overall PWID frequency and ISIDU charges did not change significantly. OD and admission frequency amongst opioid users and cost-per-patient declined between groups, suggesting a potential decrease in ISIDU and attendant costs. While median charges per admission declined, they were statistically insignificant and may represent stagnation in ISIDU-associated costs following SSP establishment. Diminishment in opioid user admissions and OD suggest additional possible positive epidemiological effects of the SSP. Weaknesses included difficulty of associating outcomes, limited post-index time period, and potential misclassification when establishing a standardized algorithm for PWID identification. Disclosures All authors: No reported disclosures.