e520 Article TypeCo respondence To the Editor:

We have read with great interest the article by DeSimone et al1 on viridans group streptococcal (VGS) infective endocarditis (IE) incidence before and after the American Heart Association (AHA) IE prophylaxis guidelines were updated in 2007. Given the lack of dem-onstrated efficacy of IE antibiotic prophylaxis, we think it of para-mount importance to identify any increase in VGS IE that would result from the restricted use of IE prophylaxis.2 We recently conducted a study that used methods different from de Simone’s (repeated popu-lation-based surveys in a large population pool of 8 million inhabit-ants) but also failed to evidence any increase in the rate of VGS IE in France since IE prophylaxis guidelines were scaled down in 2002.3 Although we concur with DeSimone’s conclusion, we have some concerns about some of their results. First, the number of VGS IE cases on which the authors concluded that the incidence of VGS IE did not increase after 2007 in Olmsted county is extremely low—3 cases over a 3-year period—and is derived from a very small popu-lation pool (≈150 000 inhabitants). Any failure to identify IE cases (resulting, for example, from patients ’ death before IE identification) may have a major impact on incidence estimation. Second, in the next part of the article, the incidence of IE, based on Nationwide Inpatient Sample (NIS) estimations and presumably extrapolated to the whole U.S. population, is not consistent with that observed in Olmsted County. Actually, considering the incidence of IE in Olmsted County during the 2007 to 2010 period (0.77 VGS IE cases per 100 000 person-years), a rough extrapolation to the 209 million U.S. adult populatio