P-1137. Decreasing Empiric Vancomycin use for Orbital Cellulitis in Children

Abstract Background Orbital cellulitis in children is often caused by Staphylococcus aureus, Streptococcus viridans group, and Streptococcus pyogenes. Oral, respiratory, and skin commensals are also considered in empiric antibiotic choice. MRSA is becoming less prevalent nationally. In 2015, our hospital implemented an evidence-based orbital cellulitis guideline to standardize care and promote antibiotic and surgical stewardship. The guidance calls for empiric clindamycin and ceftriaxone, reserving vancomycin for cases with signs of optic nerve or CNS involvement. Here we evaluate results of this guidance and summarize isolated bacteria.Fig 1.Empiric use of vancomycin pre and post guideline Methods Cases were pulled from the hospital chart by diagnosis code and positive computed tomography scan for orbital cellulitis for 4 years pre-guideline and (2009-2013) 7 years post-guideline (2015-2021) implementation and were manually reviewed for clinical diagnosis of orbital cellulitis. Variables collected included demographics, empiric antibiotics, presence of abscess, surgical intervention, bacterial identification by blood or surgical culture, length of stay, and related readmission. Statistic analysis included chi square, fisher's exact test and Wilcoxon ranked sum.Fig 2.Bacteria isolated from blood or surgical culture in all cases*13 cases were polymicrobial Results 134 cases were included in the analysis (46 pre and 88 post guideline). Median age was 7 in both groups; there was incidentally a significantly higher proportion of males in the post-guideline group (54% v 75%, p=0.025). Post guideline, empiric use of vancomycin significantly decreased from 54% to 23% (p< 0.001) (fig 1). Among cases with vancomycin use post-guidance, 60% of use was concordant with the guideline. Length of stay was reduced from a median of 4 days to 3 days (p=0.001). Rate of abscesses and surgical intervention were nearly identical in pre and post populations (63% with abscess, and 24% with surgery in both groups). Overall, bacterial cultures from blood or abscess were positive in 26% of cases (10 (22%) pre, 25 (28%) post guideline); 13 (37%) were polymicrobial. All 3 MRSA cases were susceptible to clindamycin. Conclusion Implementation of a hospital-wide guideline can result in more judicious use of broad spectrum antibiotics. Reduction in vancomycin occurred without increased length of stay, need for surgical intervention, or readmission. Disclosures All Authors: No reported disclosures