Background: Fever has traditionally been included as a minor criterion in the Duke criteria for the diagnosing infective endocarditis (IE). We evaluated the diagnostic impact of excluding fever as a minor criterion within the 2023 Duke-ISCVID (International Society for Cardiovascular Infectious Diseases) framework. Methods: : We conducted a retrospective study of suspected IE at two Swiss University Hospitals (2014-2024). The reference standard was the final diagnosis by the Endocarditis Team or expert clinicians. Each episode was classified using two versions of the 2023 ISCVID-Duke criteria: one including and one excluding fever as a minor criterion with updated definitions for definite, possible, and rejected IE. Diagnostic performance was assessed using two dichotomizations: in the first analysis, definite IE was considered positive, and possible or rejected was negative; in the second, definite or possible IE was positive, and rejected was negative. Results: IE was diagnosed in 1331/3694 (36%) episodes. Fever was present in 78% of episodes, with similar prevalence in IE and non-IE episodes (78% versus 77%; P=0.484). Using the first evaluation, sensitivity increased modestly from 74% (95% CI: 71-76%) with the original criteria to 77% (74-79%) after excluding fever and updating definitions. Using the second evaluation, specificity markedly improved from 49% (47-51%) to 80% (78-82%) with these modifications. Conclusions: Exclusion of fever as a minor criterion within the updated Duke-ISCVID framework substantially improved specificity without compromising sensitivity. These findings challenge the traditional role of fever in IE diagnosis and support a broader re-evaluation of longstanding diagnostic criteria.
Stavropoulou et al. (Wed,) studied this question.