Overall Accuracy of the Modified Duke Criteria-A Systematic Review and Meta-analysis.

Abstract

Background:  Rapid and accurate diagnosis of infective endocarditis (IE) allows timely management of this life-threatening disease and improves outcome. The Duke criteria have traditionally been the clinical method for diagnosing IE. These criteria were reformulated at different timepoints. We aimed to evaluate the real accuracy of the modified Duke criteria based on several studies that concluded the diagnosis of IE.

Methods:  Three databases were assessed. Studies were considered for inclusion if they reported the use of modified Duke criteria as the initial approach and the confirmation of the diagnosis with the gold standard methods. The meta-analysis of diagnostic test accuracy was performed after fitting the hierarchical summary receiver operating characteristic model (HSROC) with bivariate model and displaying the summarized measures of sensitivity and specificity, and positive and negative likelihood ratios.

Results:  A total of 11 studies were included. Accuracy in the included studies ranged from 62.3 to 92.2%, sensitivity ranged from 58.3 to 84.0%, and specificity ranged from 50.0 to 100%. The combined overall sensitivity and specificity were 85% (95% CI: 0.77-0.90) and 98% (95% CI: 0.89-0.99), respectively. The positive likelihood ratio was 40.2 (95% CI: 7.26-220.74) and the negative likelihood ratio was 0.15 (95% CI: 0.01-0.23).

Conclusion:  The analysis reveals that the modified Duke criteria have a high positive likelihood ratio, suggesting a robust correlation between a positive test result and the existence of IE, and a very good overall specificity at 98%. The latter aspect holds significant importance in order to prevent unnecessary overtreatment, given the intricacies involved in managing IE.