Timing of Surgery does not Affect Outcome in Patients with Infective Endocarditis - a Propensity Score Based Analysis

Abstract

Objectives: Surgery for infective endocarditis (IE) remains a challenging condition with unchanging incidence and high early and midterm mortality. Despite clear guideline recommendations, clinical decision making is hampered due to the complexity of IE patients and the optimal timing of operation is still a matter of debate. Methods: We performed a retrospective outcome analysis of 193 consecutive patients who underwent surgery for IE between 2009 and 2014 at our institution. We aimed to analyze the a) outcome after early surgery (ES) (72h; n = 171) with regard to 30-day mortality and relevant perioperative complications and b) to define the impact of the timing of surgery on mortality. To account for differences and rule out bias between compared groups, we additionally analyzed compared groups using propensity score matching for our analysis. Results: ES patients showed a higher incidence of preoperative neurologic symptoms (ES: 50% versus LS: 26,9%, p = 0,025) and septic embolism (ES: 54,5% versus LS: 30%, p = 0,021) but no significant differences with regard to the log. EuroSCORE (ES: 10.53% versus LS: 8.75%; p = 0.266). 30-day mortality was 13.6% after ES in comparison to 9.9% (p = 0.71) after LS. Concerning perioperative complications, ES patients showed a higher incidence without reaching statistical significance in the unadjusted analysis: 9.1% of patients after ES suffered from stroke (5.8% after LS; p = 0.63). The incidence of acute kidney injury (AKI) was 54.5% after ES and 47.4% after LS (p = 0.53). 18.2% of ES patients needed postoperative pacemaker implantation (PM)(LS: 12.9%; p = 0.51). After propensity-analysis (1:3 matching) neither 30-day mortality (ES: 8,3% versus LS: 16,7%, p = 0,662) nor the incidence of stroke (ES: 8,3% versus LS: 5,6%, p = 0,731), AKI (ES: 50% versus LS: 44,4%, p = 0,738) or PM-implantation (ES: 16,7% versus LS: 19,4%, p = 0,813) revealed significant differences in compared groups. Conclusion: Neither 30-day mortality nor perioperative outcome differs after early or late surgery for IE. These data might contribute to the ongoing discussion concerning the optimal timing of surgery for IE. However, decision making remains difficult and best outcome is obtained in multidisciplinary collaboration.