Endocarditis in burn patients: clinical and diagnostic considerations.

BACKGROUND: Burned patients are at high risk for invasive procedures, bacteremia, and other infectious complications. Previous publications describe high incidence, delayed diagnosis, and high mortality for endocarditis in burned patients, but do not address use of contemporary diagnostic criteria. Further analysis of the clinical presentation and diagnosis may aid in the earlier recognition and decreased mortality of endocarditis in burned patients.
METHODS: At a 40 bed burn center, during the period from 1 January 2003 to 1 August 2006, blood culture, electronic inpatient, echocardiographic, and autopsy records were reviewed for cases of endocarditis and persistent bacteremia (blood culture positivity for the same organism separated by 24h). In addition, we reviewed cases of burn-related bacterial endocarditis published in the English language. We compared the clinical and diagnostic aspects of our identified cases with those in the published literature.
RESULTS: There were 90 episodes of persistent bacteremia or fungemia in 56 of 1250 patients admitted during the study period. Echocardiography was performed on 19, identifying 4 cases of endocarditis. One additional case of endocarditis was identified post-mortem. Time until echocardiography ranged from 6 to 176 days after onset of bacteremia. Case patient age ranged from 31 to 64 years, and total burn surface area ranged from 34 to 80%. Endocarditis occurred in 0.4% of burn unit admissions and in 8.9% of these patients with persistent bacteremia. Sites involved included the mitral valve (3), tricuspid valve (2), aortic valve (1), and pulmonic valve (1). Pathogens included Staphylococcus aureus, Pseudomonas aeruginosa, and one case of Enterococcus faecium. Diagnostic clues were minimal. Case mortality was 100%. A literature review revealed 17 publications describing confirmed bacterial endocarditis in burned patients. These cases revealed a predilection for infection by S. aureus and P. aeruginosa, a relative paucity of diagnostic clues prior to death, and a trend towards ante-mortem diagnosis and increased survival with use of diagnostic echocardiography.
CONCLUSIONS: The incidence and mortality of endocarditis in burned patients remain high. Clinical clues for endocarditis in this cohort are minimal and diagnosis may be delayed. For burned patients with persistent bacteremia, especially S. aureus or P. aeruginosa of unknown source, the diagnosis of endocarditis should be entertained and early echocardiography considered.