973. Clinical Outcomes and Resource Utilization in Children Who Develop Acute Kidney Injury Following Vancomycin Use for Treatment of MRSA Bacteremia and Acute Hematogenous Osteomyelitis

Abstract Background Children with acute hematogenous osteomyelitis (AHO) and methicillin-resistant Staphylococcus aureus (MRSA) bacteremia are often treated with vancomycin (vanc) with risk of nephrotoxicity. This study evaluates resources used, and outcomes, of children with or without acute kidney injury (AKI) when vanc is used in this setting. Methods Children with AHO and MRSA bacteremia treated with vanc from 2009 to 2019 were retrospectively studied. AKI was assessed by chart review, for clinical diagnosis and by Kidney Disease Improving Global Outcomes (KDIGO) criteria. Cohorts of children with or without KDIGO AKI were compared for differences of illness severity, treatment, resource utilization, and outcomes. Multivariate logistic regression analysis was accomplished to identify factors associated with risk for AKI. Cost analysis was performed using Pediatric Health Information System (PHIS) and Healthcare Cost and Utilization Project (HCUP) databases. Results Among 631 children with AHO, 85 children with MRSA bacteremia were treated with vanc for a sufficient duration to allow trough measurements. Of these, 14 (16.5%) had chart-diagnosed AKI, whereas 24 (28.2%) children met KDIGO criteria. Children with AKI had more febrile days and higher rates of thrombosis. They had longer duration of vanc treatment (8 vs 5 d) and higher troughs (27.8 vs 17.5 mg/L). The AKI group had more blood cultures, vanc troughs, creatinine levels, and longer length of stay (LOS; 19.9 vs 11.1 days). Multivariate analysis revealed that maximum vanc trough level (odds ratio 1.05) with a cutoff of 21.7 mg/L was a significant predictor of AKI. PHIS data of 3,133 children with AHO treated with vanc identified 75 (2.4%) with AKI who had significantly longer LOS (13 vs 7 d) and higher billed charges ($117K vs $51K) than children without AKI. HCUP data of 13,223 children with AHO identified 384 (2.9%) with AKI who had longer LOS (27 vs 6 d) and higher billed charges ($317K vs $37K) than those without. Conclusion Clinical diagnosis of AKI (16.5%) grossly underestimated the KDIGO occurrence (28.2%) in our population. This study showed vanc associated AKI results in significant increases in resource utilization and healthcare costs. A high vanc trough >21.7 mg/dL is a significant contributing factor to AKI. Disclosures All Authors: No reported disclosures.