LAUSR

Infective endocarditis (IE) is a life-threatening disease and, despite its relative rarity (3–11 cases per 100 000 per year in high-income countries2,3), consumes a disproportionate share of healthcare resources, with estimated hospital costs exceeding $120 000 per patient per admission in the United States.4 Moreover, perhaps reflecting its changing demographic profile (increased frequency of nosocomial, prosthetic valve, and complex device–associated infections in an ageing population), annual mortality stubbornly exceeds 20%5,6: an outlook worse than most cancers. International consensus recognizes that improved outcomes are critically dependent on timely diagnosis and the early instigation of appropriate treatment within a specialist multidisciplinary framework.7,8 Yet, the notoriously diverse manifestations of IE and the decreasing likelihood of a classic Oslerian textbook presentation mean that clinical suspicion is often triggered after microbiological identification of potential causative organisms. The medical community is aware of the dangers of staphylococcal bacteremia,9 but appreciation of the propensity of different streptococcal species to cause IE is variable and complicated by multiple taxonomic changes and the difficulty in accurate species identification (particularly within the Viridans group of organisms). The work by Chamat-Hedemand and colleagues10 is therefore highly welcome. The authors retrospectively identified all patients with streptococcal bacteremia presenting in the capital region of Denmark over a 10-year period (2008–2017) and crosslinked these data with Danish nationwide registries to identify episodes of hospitalization associated with IE. The overall prevalence of IE among 6224 patients with streptococcal bloodstream infection was 7.1%, with wide variation among individual species (0% to 48%). Characterization according to streptococcal species proved more informative than categorization at group level, where striking variation in IE prevalence remained evident, particularly within the Mitis (Streptococcus peroris 0%, Streptococcus gordonii 44%) and Bovis (Streptococcus alactolyticus 0%, Streptococcus gallolyticus 30%) groups. In analyses adjusted for relevant clinical risk factors (age, sex, previous IE, native valve disease, presence of prosthetic material), a positive blood culture for © 2020 American Heart Association, Inc.