1020. Injection Drug Use-Associated Staphylococcus aureus Bacteremia in a Large Urban Hospital in Atlanta, Georgia

BACKGROUND: Infectious complications of injection drug use (IDU) have increased with the expanding opioid epidemic in the southeast. We assessed the incidence, clinical presentation, and treatment outcomes of IDU-associated Staphylococcus aureus (SA) bacteremia (SAB). METHODS: We created a retrospective cohort of all adults with community acquired (CA) SAB over 5 years presenting to Grady Memorial Hospital, a 1,000-bed urban county hospital in Atlanta, GA. Charts were reviewed by infectious diseases physicians to obtain clinical and laboratory characteristics, including substance use disorder (SUD), and determine if SAB was IDU-associated. The study period was divided into three periods (P1 = March 2012–January 2014, P2 = January 2014–December 2015, P3 = December 2015–November 2017) to evaluate changes in the incidence of IDU-SAB over time using Poisson regression. RESULTS: Among 321 patients with a first episode of CA-SAB, 24 (7%) were IDU-SAB. The number of IDU-SAB cases in each period increased (P1 = 4, P2 = 7, and P3 = 13 [P = 0.07 for trend]). The median age of IDU-SAB patients was 38 (IQR 31–57), 11 (46%) were black, and 15 (63%) had chronic hepatitis C virus infection. Heroin was the most common injected drug (92%) followed by cocaine (25%); multiple drugs were injected in 29%. All but two patients (92%) had a complication of SAB, most commonly endocarditis (50%) and septic pulmonary emboli (38%). The median hospitalization was 23 days (IQR 19.5–37.5) and 5 patients (12%) left the hospital against medical advice (AMA). Readmission for persistent or recurrent SA infection during the study period was common (42%), and three (13%) died ≤6 months from initial presentation, including two with prior discharge AMA. Half of the discharge summaries did not mention SUD as a hospital problem. Outpatient SUD treatment was recommended to eight (33%) patients and a recommendation of abstinence was the intervention for 12 (50%). CONCLUSION: Increasing IDU-SAB was observed over 5 years in our urban Atlanta hospital, primarily due to heroin use. Most cases were associated with complications of SAB with a long length of stay and frequent readmission, but few patients received treatment or harm reduction interventions for their SUD. These data will raise awareness and direct resources to expanding evidence-based opioid use disorder treatment for patients with infectious complications of IDU. DISCLOSURES: All authors: No reported disclosures.