The Impact of Antibiotic Escalation in Necrotizing Soft Tissue Infections: A Single Institution Experience

Authors:
Courtney Collins, Clark Ingram, Christine Dart, Anthony Gerlach, Holly Baselice, John Loftus, Jon Wisler, Anahita Jalilvand

Body of Abstract:
Introduction: Urgent surgical debridement with initial broad-spectrum antibiotics is the mainstay for achieving source control for necrotizing soft tissue infections (NSTIs). However, little is known regarding the clinical course for NSTI patients who require antibiotic escalation. This study characterizes hospital outcomes for NSTI patients requiring antibiotic escalation after debridement. Secondary objectives were determining patterns in microbial and antibiotic usage associated with escalation therapy.

Methods: We conducted a retrospective review of 625 NSTI patients at a large volume tertiary institution (2013-2023). A chart review was conducted to obtain baseline characteristics, presentation severity, and operative, microbial and antibiotic data, and 90-day outcomes. Antibiotic escalation was determined when wound culture data prompted antibiotic change due to inadequate initial therapy. In this study, we compared granular NSTI characteristics, antibiotic usage, and clinical trajectories between patients requiring antibiotic escalation (AbxEscal, 22.6%, n =141) and those who did not (n =484). A p <0.05 was considered significant. Results: Compared to controls, the AbxEscal group had higher median Charlson Comorbidity scores (4(2-6) vs 3(2-5), p = 0.05), more females (57% vs 44%, p = 0.009), and more often had a history of previous sepsis (18.4% vs 10.5%, p=0.01) and leukemia (4.3% vs 0.6%, p =0.005). Median SOFA scores were higher in the AbxEscal vs control cohort (3(1-5) vs 2(1-4), p<0.005). The AbxEscal group was more likely to have NSTIs of the head, neck, or trunk. There was no difference in time to debridement between groups. Microbial analysis revealed higher prevalence of MRSA (21% vs 11%, p = 0.006), VRE (12.1% vs 0%, p <0.005), ESBL (15.6% vs 0.8%, p <0.005), and fluconazole resistant candida (8.5% vs 0%, p <0.0005) for patients requiring antibiotic escalation. The most common escalated antibiotics were antifungals (69%), followed by Ertapenem (17%) and Daptomycin (17%). Compared to controls, the AbxEscal cohort was more likely to require ICU (81% vs 57%, p <0.005), undergo 3+ debridements (42% vs 19%, p<0.005), and exhibit higher in-hospital (12% vs 6%, p =0.05) and cumulative 90-day mortality (21% vs 11%, p =0.003). Antibiotic escalation remained an independent predictor of ICU length of stay (OR: 2.7, 95th CI: 1.6-4.6) and 3+ debridements (OR 2.9, 95th CI: 1.9-4.5) after controlling for SOFA, Charlson Comorbidity index and NSTI location. Discussion: One fifth of the cohort required antibiotic escalation based on culture data and exhibited worse outcomes. Antibiotic escalation was strongly associated with index morbidity, even after adjusting for sepsis severity and comorbidities. The correlation between antibiotic escalation and resistant microbial flora may highlight an opportunity for improving initial antibiotic management in a subset of high risk NSTI patients.