Predictors of Persistent Fever Among Patients With Suspected Infective Endocarditis: Think Outside the box.

BACKGROUND

Fever is common in infective endocarditis (IE), yet little is known about fever duration in such patients. We aim to identify predictors of persistent fever in patients with suspected IE.

METHODS

This study was conducted at the Lausanne University Hospital, Switzerland, from January 2014 to June 2023. All patients with suspected IE being febrile upon presentation were included. Fever (>38°C) was considered persistent if it continued for at least 96 hours from antimicrobial treatment initiation. A case was classified as IE by the Endocarditis Team.

RESULTS

Among 1399 episodes with suspected IE, persistent fever was observed in 260 (19%) episodes. IE was diagnosed in 536 (41%) episodes, of which 82 (15%) had persistent fever. Among episodes with suspected IE, persistent bacteremia/candidemia for 96 hours (P < .001), spondylodiscitis (P = .039), intrabdominal infection (P = .001) were associated with persistent fever. Conversely, bacteremia by streptococci (P = .049), or enterococci (P = .001), source control performed withing 96 hours (P = .015) and appropriate antimicrobial treatment within 48 hours (P = .018) were associated with early defervescence. No association between persistent fever and infective endocarditis was found (P = .207). Among 536 IE episodes, persistent bacteremia/candidemia for 96 hours (P < .001), and native bone and joint infection (P = .020) were associated with persistent fever. Conversely, bacteremia by streptococci or enterococci (P = .001; adjusted odds ratio [aOR] 0.25, 95% confidence interval [CI] .11-.58) were associated with early defervescence.

CONCLUSIONS

In episodes with suspected IE, persistent fever was associated with spondylodiscitis, inappropriate antimicrobial treatment and absence of source control interventions. Among IE patients, persistent fever was associated with native bone and joint infections.