Nearly a century ago, Dochez and Avery [1] reported their studies of the pneumococcus, which not only set the stage for modern immunology and genetics but also demonstrated that specific pneumococcal substances could be found in the urine at the onset of pneumococcal pneumonia: “a specifically reacting substance of pneumococcus origin occurs in the . . . urine of patients during lobar pneumonia . . . which reacts specifically with antipneumococcus serum of the type corresponding to the organism with which the individual is infected.” Although these “substances” were delivered to the kidney from the blood, they were concentrated in the urine and thus more easily detected there than in serum [1]. Inspired by that discovery, the urine antigen detection (UAD) method, used in the study by Sherwin et al [2] reported in this issue of The Journal, uses monoclonal antibodies to detect 13 serotype-specific Streptococcus pneumoniae polysaccharides found in the urine of patients with pneumococcal disease; the UAD test has a reported sensitivity and specificity for bacteremic pneumococcal pneumonia of 97% and 100%, respectively [3]. Tillett et al [4] further reported in 1930 that the pneumococcal C-polysaccharide was a species-specific antigen common to all pneumococci. Toward the end of the 20th century, this observation was exploited commercially in the form of the BinaxNOW urine antigen test (Alere North America), which can be used to help diagnose pneumococcal pneumonia in adults but does not provide serotype information [5]. The package-insert reported sensitivity and specificity of the BinaxNOW test for bacteremic pneumococcal pneumonia is 90% and 71%–78%, respectively. One caveat about both the UAD and BinaxNOW tests is the lack of a reference standard with which to evaluate test characteristics against nonbacteremic pneumococcal pneumonia, the real target of most investigations. Routine vaccination of US infants with a 7-valent pneumococcal conjugate vaccine (PCV7) started in 2000. The impact of PCV7 vaccination program on invasive pneumococcal disease (IPD) has been monitored through population laboratorybased surveillance. This surveillance has shown that by 2007, PCV7 serotypes were rare causes of IPD in young children, and because nasopharyngeal carriage rates of these serotypes also declined substantially, vaccine-serotype IPD had declined by >85% in older children and adults, through indirect or “herd” protection [6]. However, assessing the impact on pneumonia, a much more common pneumococcal disease, is not as straightforward, because although 30%–60% of community-acquired pneumonia in all ages has historically been attributed to the pneumococcus, a definite pneumococcal diagnosis is often not established [7]. If the pneumococcus is responsible for a large proportion of all-cause pneumonia and a substantial proportion of pneumococcal pneumonias are due to vaccine serotypes, then the large database of US hospital discharges, the Nationwide Inpatient Sample (http://www.hcup-us.ahrq. gov/nisoverview.jsp), should be able to detect changes in pneumococcal pneumonia incidence using this sensitive but nonspecific outcome. By 2004, US pneumonia hospitalizations in children aged <2 years had declined by 39% (95% confidence interval [CI], 22%–52%) relative to prePCV7 rates, and a statistically significant Received 1 July 2013; accepted 2 July 2013; electronically published 2 October 2013. Correspondence: Marie R Griffin, MD. MPH, Department of Preventive Medicine, Vanderbilt University Medical Center, Village at Vanderbilt, Ste 2600, 1500 21st Ave S, Nashville TN 37212 (marie.griffin@vanderbilt.edu). The Journal of Infectious Diseases 2013;208:1734–6 © The Author 2013. Published by Oxford University Press on behalf of the Infectious Diseases Society of America. All rights reserved. For Permissions, please e-mail: journals. permissions@oup.com. DOI: 10.1093/infdis/jit510