need for ongoing monitoring of the impact of new guidelines. Regarding the preliminary data presented at the American College of Cardiology 2014 meeting, our data on VGS were not entirely captured (only ICD9-CM codes 421.0 and 041.00 were used). Furthermore, the VGS IE diagnosis established by DeSimone et al. (2) raises serious concern because the VGS does not carry a unique ICD-9 CM code (unlike staphylococcus, entero- coccus, and so forth). This etiology was assumed by including ICD-9 CM code 041.09 or ICD-9 CM code 041.00 (“Streptococcus infection in conditions classi- fied elsewhere and of unspecified site, other Strepto- coccus”) among patients carrying the diagnosis of IE (2). The accuracy of VGS diagnosis and drawing major conclusions based on nonspecific coding can be erro- neous. Feedback from many experts attending the American College of Cardiology 2014 meeting helped us overcome this limitation in our study, resulting in elimination of VGS group. Hence, our conclusion “there has been a significant rise in the incidence of streptococcus IE following 2007 guideline” is statisti- cally sound and the study design valid. Whether the temporal association noted in our study reflects a causal relationship cannot be deduced from our study design. We acknowledged this in the limitation section. We appreciate the suggestion of DeSimone and colleagues to look at VGS as a specific subgroup. However, as pointed out by Dayer and Thornhill (5), this has to be done in a randomized controlled design to eliminate the inherent limitations of a retrospec- tive database. Sadip Pant, MD Abhishek Deshmukh, MD *Jawahar L. Mehta, MD, PhD *University of Arkansas for Medical Sciences Division of Cardiovascular Medicine 4301 West Markham Street, Mail Slot 532 Little Rock, Arkansas 72205-7199 E-mail: mehtajl@uams.edu http://dx.doi.org/10.1016/j.jacc.2015.06.1330 Please note: The authors have reported that they have no relationships relevant to the contents of this paper to disclose. REFERENCES 1. Bor DH, Woolhandler S, Nardin R, et al. Infective endocarditis in the U.S., 1998–2009: a nationwide study. PLoS One 2013;8:e60033. 2. DeSimone DC, Tleyjeh IM, Correa de Sa DD, et al. Incidence of infective endocarditis caused by viridans group streptococci before and after publica- tion of the 2007 American Heart Association’s endocarditis prevention guidelines. Circulation 2012;126:60–4. 3. Thornhill MH, Dayer MJ, Forde JM, et al. Impact of the NICE guideline recommending cessation of antibiotic prophylaxis for prevention of infective endocarditis: before and after study. BMJ 2011;342:d2392. 4. Dayer MJ, Jones S, Prendergast B, et al. Incidence of infective endocarditis in England, 2000–13: a secular trend, interrupted time-series analysis. Lancet 2015;385:1219–28. 5. Dayer M, Thornhill M. Antibiotic prophylaxis guidelines and infective endocarditis: cause for concern? J Am Coll Cardiol 2015;65:2077–8. Impact of Clinical Presentation on Dual Antiplatelet Therapy Duration Let’s Re-Evaluate Our Priorities We read with interest the paper by Yeh et al. (1) published in the Journal, reporting on a subgroup analysis from the DAPT (Dual Antiplatelet Therapy) trial in which the effect of an extended treatment with DAPT beyond 1 year was investigated in patients presenting with or without myocardial infarction at the time of stent implantation. The authors concluded that the benefit of an extended treatment persisted irrespective of the clinical presentation. We believe the results of this analysis only poorly support this conclusion for the following motivations. Reducing mortality is the ultimate goal of cardiovas- cular medicine. However, the use of combined end- points, encompassing fatal and nonfatal events, such as myocardial infarction or cerebrovascular accident, is necessary to increase study power and limit the number of patients needed in clinical trials. Death traditionally comprises a small fraction of such com- posite outcomes after percutaneous coronary inter- vention. Therefore, the underlying foundation for combining fatal and nonfatal outcomes as a reliable measure of a given treatment effect requires that nonfatal endpoints are independently associated with fatal events and that the strength of this asso- ciation is somewhat comparable across nonfatal endpoints. The rationale for extending DAPT beyond the recommended period is to prevent myocardial infarction, both stent- and nonstent-related, and by that improving survival. In contrast, the DAPT study showed an increase in mortality by an extended course of treatment, and the subgroup analysis by Yeh et al. (1) strongly suggests that the excess of fatality originates from patients presenting without myocardial infarction. In this patient subset, prolonged DAPT duration was associated to a 43% mortality increase. However, the interaction testing for mortality did not reach the formal level of significance (P int : 0.13). Importantly, interaction testing is known to be underpowered, and JACC VOL. 66, NO. 10, 2015 Letters SEPTEMBER 8, 2015:1198 – 205 1203