Abstract 11890: Timing of Surgery for Infective Endocarditis: A Network Meta-Analysis

Introduction: Infective endocarditis (IE) is associated with in-hospital mortality of 15-30%. Surgery is recommended in approximately 50%. The best available evidence from meta-analyses has demonstrated benefit of early surgical intervention (SI) for the management of IE. However, the timing of surgery has not been further sub-stratified and internally compared Methods: We performed a frequentist network meta-analysis adhering to PRISMA Guidelines. Observational studies comparing SI for infective endocarditis at different time points - SI performed <48 hours of diagnosis, 2-7 days, 7-14 days, 14-28 days, 28+ days - were included for analysis. In-hospital mortality, all-cause mortality, in-hospital embolic events, and recurrence of endocarditis were used as outcome variables. All analysis was performed on R, using the open-source net meta-package. Results: A total of 28 studies met our inclusion criteria. Figure 1A shows forest plots for each outcome. SI at any time period improved all-cause mortality. However, only Surgery > 21 days resulted in improved in-hospital mortality (OR = 0.19 (0.04-0.85)). Intervention at any specific surgical window did not improve in-hospital embolic events or recurrence of endocarditis. Figure 1B shows SUCRA-dependent hierarchical ranking. Surgery greater than 21 days (S = 0.9443), followed by Surgery < 48 hours (S = 0.7133) was superior to other intervention windows. Surgery < 48 hours also was found to have a higher rank for in-hospital embolic events (S = 0.947). Conclusions: Our results supplement the existing society consensus statements and guidelines to opt for early SI, by positing the value of SI within 48 hours. If the patient is stable SI can be performed after 21 days for better in hospital mortality rates. Randomized control trials need to be performed to evaluate further.