Predictors of mortality in early surgical intervention for active native valve endocarditis and significance of antimicrobial therapy: A single-center experience

Background and aim of the study: Cardiac surgery for active infective endocarditis remains a challenging and high-risk procedure. The outcome from early surgical intervention for active native valve endocarditis (ANVE) was studied, the aim being to identify significant predictors of mortality and the relationship between duration of preoperative antibiotics and outcome. Methods: Between January 1996 and February 2002, 61 patients with ANVE underwent surgery within four weeks of diagnosis. Preoperatively, 29 patients received antibiotics for < 2 weeks (group A), and 32 received antibiotics for 2-4 weeks (group B). The median follow up period was 37.4 months (range: 2155 months). Data were collected retrospectively and analyzed. To determine factors related to mortality, Kaplan-Meier survival analysis was employed, utilizing log-rank statistics to identify evidence of significant differences between the groups. The relationship between the duration of preoperative antibiotics and morbidity was determined using chi-square and Fisher's Exact tests, as appropriate. Results: Overall operative mortality was 14.8% (group A, 13.8%; group B, 15.6%). Rates of early and late prosthetic valve endocarditis were 1.8% and 1.9% (only in group B) respectively. The overall survival rate for the follow up period was 81.9%. Predictors of mortality were extensive infection (p = 0.01), poor left ventricular function (p < 0.0001), progressive cardiac failure as an indication for surgery (p < 0.0001), postoperative sepsis (p < 0.0001), renal failure after surgery (p = 0.0002) and use of a bioprosthetic valve (p = 0.045). There were no significant inter-group differences for extensive infection (p 1.00), postoperative sepsis (p = 1.00), reoperation (p 1.00) and mortality (p = 1.00). Conclusion: In patients with ANVE, early aggressive surgical intervention before the onset of cardiac failure and spread of infection is warranted. The present data suggest that, in these patient groups, the duration of preoperative antibiotics had no significant influence on postoperative morbidity and mortality.