Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Estimating the Burden of Pneumococcal Pneumonia among Adults: A Systematic Review and Meta-Analysis of Diagnostic Techniques

Estimating the Burden of Pneumococcal Pneumonia among Adults: A Systematic Review and Meta-Analysis of Diagnostic Techniques

  • Maria A. Said, 
  • Hope L. Johnson, 
  • Bareng A. S. Nonyane, 
  • Maria Deloria-Knoll, 
  • Katherine L. O′Brien, 
  • for the AGEDD Adult Pneumococcal Burden Study Team
PLOS
x

Abstract

Background

Pneumococcal pneumonia causes significant morbidity and mortality among adults. Given limitations of diagnostic tests for non-bacteremic pneumococcal pneumonia, most studies report the incidence of bacteremic or invasive pneumococcal disease (IPD), and thus, grossly underestimate the pneumococcal pneumonia burden. We aimed to develop a conceptual and quantitative strategy to estimate the non-bacteremic disease burden among adults with community-acquired pneumonia (CAP) using systematic study methods and the availability of a urine antigen assay.

Methods and Findings

We performed a systematic literature review of studies providing information on the relative yield of various diagnostic assays (BinaxNOW® S. pneumoniae urine antigen test (UAT) with blood and/or sputum culture) in diagnosing pneumococcal pneumonia. We estimated the proportion of pneumococcal pneumonia that is bacteremic, the proportion of CAP attributable to pneumococcus, and the additional contribution of the Binax UAT beyond conventional diagnostic techniques, using random effects meta-analytic methods and bootstrapping. We included 35 studies in the analysis, predominantly from developed countries. The estimated proportion of pneumococcal pneumonia that is bacteremic was 24.8% (95% CI: 21.3%, 28.9%). The estimated proportion of CAP attributable to pneumococcus was 27.3% (95% CI: 23.9%, 31.1%). The Binax UAT diagnosed an additional 11.4% (95% CI: 9.6, 13.6%) of CAP beyond conventional techniques. We were limited by the fact that not all patients underwent all diagnostic tests and by the sensitivity and specificity of the diagnostic tests themselves. We address these resulting biases and provide a range of plausible values in order to estimate the burden of pneumococcal pneumonia among adults.

Conclusions

Estimating the adult burden of pneumococcal disease from bacteremic pneumococcal pneumonia data alone significantly underestimates the true burden of disease in adults. For every case of bacteremic pneumococcal pneumonia, we estimate that there are at least 3 additional cases of non-bacteremic pneumococcal pneumonia.

Introduction

Community-acquired pneumonia (CAP) is a significant cause of morbidity and mortality among adults worldwide, of which a significant proportion is believed to be caused by Streptococcus pneumoniae (pneumococcus). The World Health Organization (WHO) estimated in 2005 that among all age groups, pneumococcal disease caused an estimated annual 1.6 million deaths [1]. Although the global pneumococcal disease burden among children is well understood, with an estimated 13.9 million cases of pneumococcal pneumonia occurring among children <5 years of age in 2000 [2], [3], the pneumococcal pneumonia burden among adults is not well characterized, impeding policy formulation for prevention and treatment.

In the pre-antibiotic era, when organisms were identified by culture and mouse inoculation, 95% of lobar pneumonia cases [4] were attributed to pneumococcus. S. pneumoniae is still thought to be the most common etiologic agent of CAP [5], [6], [7], [8] but is now identified in a much lower proportion of patients than historically.This change is thought due to a decrease in microbiological testing and the use of empiric antibiotics prior to testing, which has increased the proportion of cases with unknown etiology [9]. Many of these cases are believed to be attributable to pneumococcus [10].

Our objective was to develop a conceptual and quantitative strategy to estimate the burden of non-bacteremic pneumococcal pneumonia among adults by assessing the yield of blood cultures, sputum cultures, and the Binax UAT to establish the proportion of pneumococcal pneumonia that is bacteremic and the proportion of CAP attributable to pneumococcus (Figure 1) (See Appendix S1: Said, MA, et al. Estimating the burden of pneumococcal pneumonia among adults: the conceptual framework). Here, we present estimates for these values, achieved through a systematic review and meta-analysis of studies that report results of the Binax UAT as well as blood culture and/or sputum culture. We also present an estimate of the additional contribution of the Binax UAT to the conventional diagnostic techniques of blood and sputum culture.

thumbnail
Figure 1. The relationships between CAP1, non-bacteremic pneumococcal pneumonia, and bacteremic pneumococcal pneumonia.

https://doi.org/10.1371/journal.pone.0060273.g001

Methods

Search Strategy

A conceptual framework and modeling strategy was first developed (see Appendix S1). The model was used to guide this systematic review and meta-analysis, which was prepared in accordance with guidelines for meta-analyses of observational studies [11] as well as the PRISMA statement [12]. A previously existing review protocol for this particular study was not known to exist. We performed three separate literature searches in Medline, Embase, CINAHL, Global Health, and ISI Web of Knowledge without date restrictions to identify papers reporting the diagnostic yield of the BinaxNOW® S. pneumoniae urine antigen test (UAT) as well as blood culture and/or sputum culture in cases of CAP among adults and that would allow for a comparison between the Binax UAT and at least one of the other two tests. The literature search was performed by one author (MS) in May–June 2010 with the assistance of a medical librarian. Titles and abstracts were screened for potentially relevant citations and the full-text was retrieved for studies reporting the diagnostic yield of the BinaxNOW® S. pneumoniae urine antigen test (UAT) as well as blood culture and/or sputum culture in cases of CAP among adults. The first search aimed to identify papers that reported the usefulness of the Binax UAT in diagnosing pneumococcal pneumonia; in Pubmed, the following search terms were used: (("binax"[all fields] OR "binaxnow" OR "urinary antigen test")) AND ("Streptococcus pneumoniae"[Mesh] OR "Pneumococcal Infections"[Mesh] OR "streptococcus pneumoniae"[all fields] OR "pneumococcal"[all fields] OR "diplococcus pneumoniae"[all fields] OR "pneumococcus"[all fields] OR "pneumococci"[all fields] OR "S. pneumoniae"[all fields] OR "pneumococcal infection"). The second search aimed to identify papers that reported the usefulness of the Binax urinary antigen test in all cases of pneumonia; search terms used were: (("Pneumonia"[Mesh] OR "pneumonia")) AND ("binax" OR "binaxnow" OR "urine antigen" OR "urinary antigen"). The third search broadened the terms used to identify studies using the Binax UAT by including the search terms "streptococcus pneumoniae antigen" OR "urinary pneumococcal antigen". References lists of obtained papers as well as suggestions from other authors were used to identify additional studies.

A study was eligible for inclusion if it reported primary data from human studies of radiologically confirmed adult pneumonia cases. We considered all studies that reported results of the Binax UAT as well as blood culture and/or sputum culture. We excluded studies that looked at particular patient populations that were not representative of the general adult CAP population; for example, a study conducted among patients with malignancy. We chose not to exclude a single paper in which all patients were infected with HIV, as we were particularly interested in how HIV infection might affect the outcome measures. Studies about cases already identified as pneumococcal were excluded in order to reduce bias. We included studies utilizing the Binax UAT on both concentrated and non-concentrated urine. Sputum culture results were accepted only if sputum quality criteria were used. All foreign language articles were translated, either through electronic or official translation services. Study population characteristics, case ascertainment and diagnostic methods, and relevant outcome data were abstracted from studies after full-text screening. We also attempted to contact the authors of potentially relevant studies to obtain additional information, including the number of patients undergoing each diagnostic test, the number of positive test results, and the overlap in positive results among the three methods for detection (Figure 2).

thumbnail
Figure 2. The relationship in diagnostic test yield of blood culture, sputum culture, and the Binax UAT2.

https://doi.org/10.1371/journal.pone.0060273.g002

Pneumococcal pneumonia was defined as radiographic pneumonia in addition to at least one positive laboratory test: a blood culture or sputum culture positive for pneumococcus or a positive Binax UAT. Bacteremic pneumococcal pneumonia was defined as having a blood culture positive for pneumococcus, and non-bacteremic was defined as having a sputum culture positive for pneumococcus and/or a positive Binax UAT without a blood culture positive for pneumococcus.

Study Selection

Of the 488 articles originally identified, 140 were chosen for full text screening. Of these, 75 were identified to have potentially useful information on comparative diagnostic yield of the tests; authors were contacted to provide additional information and clarification. When a subset of the study population contained a population of CAP patients thought appropriate for inclusion, (i.e. a study of patients with lower respiratory tract infections among whom some were identified as having radiologically confirmed CAP), the author was asked to submit data on the desired subset. When an abstract without a published paper was identified that was thought to represent a study that might include the needed data, the author was also asked for additional data. Authors known to have worked or be working on a study that was thought to be suitable but had not yet been published were also contacted. Contact was conducted over email; if no email address for a contacting author was identifiable, a letter was sent by mail. Several publications reported results from the same cohort of CAP patients and were classified as one study. Contact was made with a representative from 49 of the 75 papers; a completed abstraction form was returned for 33 study cohorts, not of all which were found to meet criteria to be included in the study. Six of the studies included in the final analysis were included based on only the information provided in the published paper, as we were unable to communicate with the author. Thirty-five studies were ultimately included in the analysis.

Data extraction

Extracted data included the country in which the study was conducted, the time period over which the study took place, the language in which the paper was published, the nature of the study population, the presence or absence of a chest x-ray requirement in the diagnosis of pneumonia, the presence or absence of criteria by which to judge sputum quality, the mean age of participants, the percent of participants that were hospitalized, the percent of patients in an ICU, the mean PSI score and the distribution of participants among the five PSI classes, the percent of participants vaccinated against pneumococcus, the percent of patients with HIV, the percent of patients who came from a nursing home, the percent of patients who received antibiotics before diagnostic testing, the numbers of patients who underwent each diagnostic test, whether concentrated or unconcentrated urine was used to perform the Binax UAT, the numbers of study participants with positive blood cultures, sputum cultures, and the Binax UAT and the overlap in positivity of these test results, the numbers of participants with positive test results for each of the diagnostic tests studied stratified by whether they had received antibiotics prior to diagnostic testing, and the number of positive test results for each of the diagnostic tests studied stratified by the Pneumonia Severity Index (PSI) class or CURB criteria. Data were extracted and classified by one author (MS) and entered into a Microsoft Access 2007 database.

Statistical Analysis

Studies in which only two diagnostic tests were used were weighted the same as those studies in which three diagnostic tests were used, but these studies did not contribute to all proportions calculated. In cases in which zero positive cases were identified, we added 0.5 to the numerator and 1.0 to the denominator, as study weights (the inverse of the variance) would otherwise be undefined.

Studies were stratified by the proportion of study participants who received antecedent antibiotics into high and low categories. This proportion was not known for three studies. In two studies from Spain, the proportion was assumed to be the mean of the proportion of other studies from Spain. In a study among Navajo, the proportion was assumed to be low, and this study was categorized with other studies in which the proportion on antibiotics was low.

Severity of illness was defined using the Pneumonia Severity Index (PSI), which uses an algorithm to identify patients with CAP at low risk of dying within 30 days of presentation: a score of 1 represents the lowest risk and a score of 5 represents the highest risk [13]. In some cases, studies that did not record the patients' PSI class were characterized as having severe or non-severe populations based on the proportion in the ICU or the proportion hospitalized.

The methods and estimates were developed through an interactive process that included discussions with an independent Expert Review Panel as well as independent consultations with additional experts on CAP and pneumococcal pneumonia. Revisions were based on their comments and suggestions.

We sought to determine the proportion of pneumococcal pneumonia that is bacteremic, the proportion of CAP attributable to pneumococcus, and the contribution of the Binax UAT in diagnosing pneumococcal pneumonia over and beyond conventional culture techniques. Random effects meta-analysis to summarize the results was used because significant heterogeneity across studies was observed [14]. Analysis was done on the log-transformed study estimates. The numbers of study participants who underwent blood cultures, sputum cultures, or the Binax UAT were known; however, the numbers of people who underwent more than one diagnostic test were not (we knew that x people underwent blood culture and y people underwent sputum culture but not if the person who underwent blood culture was part of the group who underwent sputum culture). Thus, to calculate the proportion of CAP identified as pneumococcal, we used the total study population as the denominator, assuming all subjects were tested by all methods. We then evaluated the proportion of CAP identified as pneumococcus by blood culture, sputum culture, or the Binax UAT among those who actually underwent each test.

Sensitivity analysis was done by excluding outlying study values in order to determine their influence. We also stratified studies by proportion of pneumococcal pneumonia cases who were HIV-infected (<20% or unknown vs. ≥20%), severity of illness, and prior antibiotic use. Univariate linear regression (which assigns equal weights to studies) and random effects meta-regression were used to explore potential confounding factors for the proportion of CAP identified as pneumococcus. The Monte Carlo permutation test was applied to the meta-regression to obtain p-values adjusted for multiple testing [15]. A significance level of 0.05 was used to identify potential confounders between studies, which included hospital admission, HIV infection, antibiotics received before diagnostic testing, mean age, mean Pneumonia Severity Index (PSI), and whether concentrated urine was used for the Binax UAT testing. Finally, we analyzed the effect of (a) receipt of antibiotics before specimen collection and (b) PSI class comparing higher (IV–V) vs. lower (I–III) PSI classes on diagnostic yield. Heterogeneity of studies was evaluated by the I-squared test. Statistical analyses were done in Stata (version 11) [16].

Results

Review of citations' titles and abstracts identified from a systematic literature search and hand searching of citation lists from other relevant published studies yielded 488 articles (Figure 3). Thirty-five studies reported the yield of the Binax UAT relative to blood culture and/or sputum culture in diagnosing adult pneumococcal pneumonia and were included in the analysis; of these, 28 had data for the relationship of positive results for all three diagnostic tests and were used in the calculations of the proportion of pneumococcal pneumonia that is bacteremic and the proportion of CAP attributable to pneumococcus. The included studies represented data from 18 countries (Table 1), predominantly from Europe and North America. The majority of included studies were prospective studies of CAP etiology among hospitalized cases.

The proportion of pneumococcal pneumonia that was bacteremic ranged from 2.2% to 50.9% (median 28.9%, IQR 14.8% to 33.4%). In the meta-analysis, the proportion of pneumococcal pneumonia estimated to be bacteremic was 24.8% (95% CI: 21.3%, 28.9%) (Figure 4). The proportion of CAP identified as pneumococcus varied by diagnostic test. It was smallest with blood cultures and greatest with the Binax UAT (Figure 5). The meta-estimate of the proportion of CAP attributable to pneumococcus was 27.3% (95% CI: 23.9%, 31.1%). Inclusion criteria were used to maximize the quality of the studies included in the analysis. However, studies with larger population sizes were thought to provide more reliable proportions and were analyzed separately in order to assess this particular bias. When studies with <200 participants were compared with studies with ≥200 participants, the proportion of CAP attributable to pneumococcus was 27.4% (95% CI: 22.2%, 33.9%) in the smaller studies compared to 26.7% (95% CI: 22.7%, 31.4%) in the larger studies.

thumbnail
Figure 4. Forest plot for the proportion of pneumococcal pneumonia identified as bacteremic, sorted by proportion of study participants who received antecedent antibiotics.

https://doi.org/10.1371/journal.pone.0060273.g004

thumbnail
Figure 5. The proportion of community-acquired pneumonia attributable to pneumococcus, according to diagnostic test results, by individual studies3.

https://doi.org/10.1371/journal.pone.0060273.g005

When we used as the denominator the number of people who actually underwent a diagnostic test rather than the total study sample size to evaluate the pneumococcal yield of each individual test, the proportion of CAP diagnosed as pneumococcal rose from 7.5% (95% CI: 6.4%, 8.8%) to 8.1% (95% CI: 7.0%, 9.5%) by blood culture, from 9.8% (95% CI: 8.1%, 12.0%) to 19.9% (95% CI: 16.9%, 23.4%) by sputum culture, and from 20.0% (95% CI: 17.4%, 23.1%) to 23.9% (95% CI: 21.7%, 26.4%) by the Binax UAT.

The Binax UAT increased the diagnostic yield over and above blood and sputum culture by an additional 11.4% (95% CI: 9.6%, 13.6%). Expressed as a ratio, the Binax UAT identified an additional 0.87 cases (95% CI: 0.51, 1.36) for every case identified as pneumococcal by blood and sputum culture.

We also stratified the results by prior antibiotic use, disease severity, and HIV prevalence (Table 2). In meta-regression, none of the included potential confounders were found to be significantly associated with proportion estimates. There was a high degree of heterogeneity in all outcomes of interest (e.g. I-squared for the proportion of CAP identified as pneumococcal by all three diagnostic tests was 91.6%), and none of the potential confounders included in the meta-regression explained this heterogeneity (adjusted p-values >0.05 for all covariates).

thumbnail
Table 2. Meta-analysis of the proportion of pneumococcal pneumonia that is bacteremic and the proportion of CAP attributable to pneumococcus, stratified by severity of disease, prior antibiotic use, and HIV status.

https://doi.org/10.1371/journal.pone.0060273.t002

We further investigated the effects of prior antibiotic use and PSI class (IV–V vs. I–III) on the diagnostic yield of each test and calculated the risk ratio for having a positive blood culture, sputum culture, or Binax UAT (Table 3). Prior antibiotics reduced the relative diagnostic yield for blood cultures by 67% (95% CI: 53%, 77%), for sputum cultures by 34% (95% CI: 8%, 53%) and for the Binax UAT by 26% (95% CI: 0%, 44%). The relative pneumococcal diagnostic yield among those people with more severe disease (PSI class IV–V) compared to those with less severe disease (PSI class I–III) increased for blood cultures by 72% (95% CI: 38%, 115%), and for the Binax UAT by 31% (95% CI: 15%, 50%).

thumbnail
Table 3. Meta-analysis of the association of prior antibiotic use and PSI class on the yield of blood culture, sputum culture, and the Binax UAT.

https://doi.org/10.1371/journal.pone.0060273.t003

Discussion

Through evaluation of the relationship in diagnostic yield of blood culture, sputum culture, and the Binax UAT in diagnosing adult pneumococcal pneumonia, we estimated the proportion of pneumococcal pneumonia that is bacteremic to be approximately 25% and, thus, the ratio of non-bacteremic pneumococcal pneumonia to bacteremic pneumococcal pneumonia to be approximately 3∶1. Because this estimate assumed a normal distribution, we also performed a bootstrap analysis [17] to determine the ratio of non-bacteremic pneumococcal pneumonia to bacteremic pneumococcal pneumonia, which gave a similar result. The wide range among studies in the proportion of pneumococcal pneumonia that was bacteremic could not be explained by study size, study location, proportion of participants undergoing sputum cultures and Binax UAT testing, proportion of participants on antecedent antibiotics, and severity of illness; however, in the two studies with the lowest proportion of bacteremic cases, many people did not undergo blood cultures and thus, some bacteremic cases were probably not identified. In the study with the most extreme value for this proportion (2.2%), only 11 of 80 study participants underwent blood culture testing, likely leading to a proportion underestimate.

This finding of a ratio of non-bacteremic to bacteremic pneumococcal pneumonia of 3∶1 is consistent with estimates from the pre-antibiotic era, which suggested that for every case of bacteremic pneumococcal pneumonia, there were 2–4 cases of non-bacteremic pneumococcal pneumonia [18], [19], [20]. Diagnosis at that time relied on extensive microbiological techniques and was unaffected by antibiotic use. Although our estimates are similar to those from the pre-antibiotic era, we used results from only three diagnostic tests, and the test with the highest yield, the Binax UAT, is known in cases of non-bacteremic pneumococcal pneumonia to have a sensitivity of as low as 52% among patients with sputum cultures positive for pneumococcus [21]. Furthermore, many patients in these etiologic studies received antecedent antibiotics, and not all patients underwent every diagnostic test. Thus, we believe that our estimate of the proportion of pneumococcal pneumonia that is bacteremic represents a lower limit of the true proportion.

We found that at least one-quarter of CAP cases are likely attributable to pneumococcus. This is consistent with a systematic review of 127 study cohorts among adults from 1966–1995 that found that pneumococcus accounted for 24% of all pneumonia cases [22]. Another source of validation is a study analyzing the impact of infant pneumococcal conjugate vaccination on U.S. pneumonia and influenza hospitalization and mortality rates for all ages. The study estimated, using modeling techniques, a reduction in the proportion of CAP attributable to pneumococcus from about 35–40% in the pre-vaccine era to about 18–28% in the post-vaccine era [23]. As the Binax UAT test was introduced around the same time as the pediatric pneumococcal conjugate vaccine, it is possible that the results from our analysis may reflect some indirect effects of the vaccine. Our estimate of the proportion of CAP attributable to pneumococcus is somewhat lower than the estimate for children; using a vaccine-probe approach, a meta-analysis of 4 randomized pediatric pneumococcal conjugate vaccine efficacy trials estimated the proportion of radiologically confirmed pneumonia attributable to pneumococcus to be 36% [2]. However, some of the pediatric vaccine trials were conducted in resource-limited countries and may explain the higher pneumococcal pneumonia burden compared to our findings among adults in high-resource settings.

All of the studies used to estimate the proportion of pneumococcal pneumonia that is bacteremic and the proportion of CAP attributable to pneumococcus by all 3 diagnostic methods were performed in resource-rich countries; the study from Nicaragua did not include data on the yield of blood cultures and, thus, did not contribute to these estimates. Given true differences in pneumonia epidemiology in resource-limited compared to resource-rich populations, such as age distribution, prevalence of nasopharyngeal pneumococcal colonization, the prevalence of underlying illness (e.g. HIV), and the ability to identify the etiology of CAP due to disparities in laboratory testing, access to care, and antecedent antibiotic use, it is possible that both the true and estimated proportion of pneumococcal pneumonia that is bacteremic and the proportion of CAP attributable to pneumococcus might differ. The paucity of data from resource-limited settings, particularly Africa, makes these differences difficult to assess. One recent study from Kenya, which used a serotype-specific latex agglutination urine antigen test, estimated by latent class analysis the proportion of CAP attributable to pneumococcus in adults to be 46% (95% CI: 36%–57%), over 18% higher than what we found [24].

Etiology studies among adults in resource-limited settings often use blood culture and sputum examination. By estimating the additional contribution of the Binax UAT in diagnosing pneumococcal pneumonia, we demonstrate that the Binax UAT may identify an additional 11% with an etiology of pneumococcal pneumonia compared to studies that only use blood culture and sputum examination. A recent study in South Africa of a rapid molecular assay for nasopharyngeal pneumococcal density in a population with high HIV prevalence found, similar to this study, that blood culture, good quality sputum culture and gram stain identified 15.4% of the population with pneumococcus and that the Binax UAT added an additional 11.7% [25].

The majority of studies included only hospitalized patients; only 5 studies included non-hospitalized patients. When we excluded these studies from the analysis, the results did not change meaningfully. The proportion of pneumococcal pneumonia that was bacteremic demonstrated an increasing trend in studies with more severe disease, which is consistent with findings that bacteremia is more common among patients with severe disease [26]. The proportion of pneumococcal pneumonia that was bacteremic was lower, although not statistically significantly lower, in studies in which a high proportion of the study population received antecedent antibiotics, likely indicating the greater effect antibiotic use has on blood culture yield compared to other diagnostic tests. The majority of studies included no patients with known HIV infection. Among three studies with an HIV prevalence ≥20%, a greater proportion of CAP was attributable to pneumococcus than among the other studies. This finding is consistent with studies that reported greater risk of developing IPD or bacterial pneumonia among HIV-infected individuals compared to the general population [27], [28].

We were able to further investigate, through a meta-analysis of risk ratios, the effect of antibiotic use and severity of illness on pneumococcal diagnostic test positivity. As expected, we found a decreased yield of blood culture and sputum culture among those who had received previous antibiotics. We also found a decreased yield of the Binax UAT after antibiotics, although the result was not statistically significant. Previous studies have shown conflicting results as to whether antecedent antibiotic use decreases Binax UAT yield [29], [30], [31]. We found increased yield of the three diagnostic tests with increased severity of disease, although for sputum culture this was not statistically significant. Many studies have shown that conventional diagnostic tests of blood culture and sputum culture have higher yield among sicker patients [26], [32], [33]. Studies among adults have demonstrated the Binax UAT to have a sensitivity of 77%–92% [21], [29], [34], [35], [36] in diagnosing bacteremic pneumococcal pneumonia and 52%–78% in diagnosing non-bacteremic pneumococcal pneumonia [21], [29], [34], [35], suggesting that the UAT is more sensitive in bacteremic patients; however the studies are difficult to compare as each defines non-bacteremic pneumococcal pneumonia differently. In some, but not all studies, the Binax UAT yield increased with higher PSI score [29], [30], [32].

There are a number of limitations to our study. Most notably, our findings are best generalizable to CAP patients presenting to hospital in developed countries. We only included studies in which the Binax UAT was used and thus were not able to include many studies from resource-limited countries. The direction of this bias depends on multiple factors and is difficult to predict. Higher HIV prevalence in some settings might lead to a higher true proportion of CAP attributable to pneumococcus, but a lack of adequate lab facilities in resource-limited settings might reduce the diagnostic yield and thus suggest a lower reported proportion of CAP attributable to pneumococcus. The studies included may also not be representative of the general CAP population, as almost all the patients were hospitalized and thus had moderate to severe disease and/or likely increased prevalence of comorbidities. An outpatient population with less severe disease would be expected to have a smaller proportion of pneumococcal pneumonia that was bacteremic than a hospitalized patient population, and thus, our estimate might underestimate the true proportion. By estimating the additional contribution of the Binax UAT over the conventional diagnostic tests (blood and sputum culture) and by estimating the degree to which antecedent antibiotic use and severity of illness is associated with diagnostic test yield, we aimed to provide a range of plausible estimates. For example, although the estimates here are representative of a hospitalized population in a resource-rich setting, the finding that the Binax UAT identified an additional 0.87 cases (95% CI: 0.51, 1.36) for every case identified as pneumococcal by blood or sputum culture could be used when estimating the burden of pneumococcal pneumonia in populations in which the Binax UAT was not used.

We were also limited by the fact that not all patients underwent all diagnostic tests. Blood cultures were performed on 14%–100% (median 97% and mean 89%) of participants. Sputum cultures were performed on 20%–100% of participants (median 60% and mean 60%); and the Binax UAT was performed on 6%–100% (median 96% and mean 90%) of participants. By assuming that all participants underwent each test, we underestimated the proportion of people infected with pneumococcus. Because sputum cultures were performed on fewer people than blood cultures, when we calculated the proportion of CAP cases diagnosed as pneumococcal by each diagnostic test using the true number of study participants who underwent each test, rather than all study participants, the yield of sputum culture rose significantly; thus, this limitation may have led to an overestimation of the proportion of pneumococcal pneumonia that is bacteremic. Furthermore, we do not know whether those people who underwent certain diagnostic tests were qualitatively different than those who did not undergo certain diagnostic tests. If blood cultures were collected more often by some clinicians than others, but urine was collected on all subjects, this might bias our results to underestimate the proportion of pneumococcal pneumonia that is bacteremic.

Finally, our results are constrained by the limitations in sensitivity and specificity of the diagnostic tests themselves. For example, assuming that the Binax UAT is more sensitive among bacteremic patients, it might miss a larger number of non-bacteremic patients and lead to a falsely high proportion of pneumococcal pneumonia thought to be bacteremic.

Our study shows that estimates of the number of pneumococcal pneumonia adult cases that rely primarily on blood culture results vastly underestimate the true burden of disease. The objective of this analysis was not to conduct a meta-analysis of the tests' performances but to provide a range of plausible values in order to estimate, with tests known to be imperfect, the burden of non-bacteremic pneumococcal pneumonia among adults. Additional studies utilizing the Binax UAT and development and use of more reliable tests for non-bacteremic pneumococcal pneumonia, including possibly quantitative sputum or nasopharyngeal PCR, will improve our understanding. The observed reduction in both IPD and pneumococcal pneumonia among unvaccinated older children and adults due to the indirect effects of pediatric conjugate vaccine can also provide a measure of pneumococcal pneumonia burden [23], [37], [38]. Through this and additional studies, a better understanding of the adult disease burden can help guide treatment and prevention strategies.

Supporting Information

Figure S1.

The relationship between diagnostic testing and the proportion of CAP attributable to pneumococcus.

https://doi.org/10.1371/journal.pone.0060273.s001

(EPS)

Appendix S1.

Estimating the burden of pneumococcal pneumonia among adults: the conceptual framework.

https://doi.org/10.1371/journal.pone.0060273.s002

(DOCX)

Acknowledgments

AGEDD Adult Pneumococcal Burden Study Team Group Co-authors:

• Felipe Andreo, MD, PhD, Servei de Pneumologia, Hospital Univ. Germans Trias i Pujol, Institut d'Investigació Germans Trias i Pujol, Universitat Autònoma de Barcelona, CIBER Enfermedades Respiratorias, Badalona, Barcelona, Spain

• Bojana Beovic, MD, PhD, Department of Infectious Diseases, University Medical Centre Ljubljana, Ljubljana, Slovenia

• Silvia Blanco, MSc, PhD, Servei de Microbiologia, Hospital Univ. Germans Trias i Pujol, Institut d'Investigació Germans Trias i Pujol, Universitat Autònoma de Barcelona, CIBER Enfermedades Respiratorias, Badalona, Barcelona, Spain

• Wim G. Boersma, MD, PHD, Department of Pulmonary Diseases, Alkmaar, The Netherlands

• David R. Boulware MD, MPH, CTropMed, Department of Medicine, University of Minnesota, Minneapolis, MN, USA

• Jay C. Butler, MD, Alaska Native Tribal Health Consortium, Anchorage, AK, USA

• Jordi Carratalà, MD, PhD, Department of Infectious Diseases, Hospital Universitari de Bellvitge-IDIBELL, University of Barcelona, L`Hospitalet de Llobregat, Barcelona, Spain

• Feng-Yee Chang, MD, PhD, Department of Internal Medicine, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan

• Patrick G. P. Charles, MBBS, PhD, FRACP, Department of Infectious Diseases, Austin Health, Heidelberg, Australia

• Alejandro A. Diaz, MD, Pulmonary and Critical Care Division, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA and Department of Pulmonary Diseases, Pontificia Universidad Católica de Chile, Santiago, Chile

• Jose Domínguez, MSc, PhD, Servei de Microbiologia, Hospital Univ. Germans Trias i Pujol, Institut d'Investigació Germans Trias i Pujol, Universitat Autònoma de Barcelona, CIBER Enfermedades Respiratorias, Badalona, Barcelona. Spain

• Naomi Ehara, MD, PhD, Division of Respiratory Medicine, Japanese Red Cross Nagasaki Genbaku Isahaya Hospital, Nagasaki, Japan

• Henrik Endeman, MD, PhD, Department of Intensive Care Medicine, Onze Lieve Vrouwe Gasthuis, Amsterdam Netherlands

• Vicenç Falcó, MD, Department of Infectious Diseases, Hospital Universitari Vall d'Hebron, Universitat Autònoma de Barcelona, Barcelona, Spain

• Miquel Falguera, MD, Department of Internal Medicine, Hospital Universitari Arnau de Vilanova, Lleida, Spain

• Kiyoyasu Fukushima, MD, PhD, Division of Respiratory Medicine, Japanese Red Cross Nagasaki Genbaku Isahaya Hospital, Nagasaki, Japan

• Carolina Garcia-Vidal, MD, PhD, Department of Infectious Diseases, Hospital Universitari de Bellvitge-IDIBELL, L'Hospitalet de LLobregat, Barcelona, Spain.

• Daniel Genne, MD, Hospital Neuchatelois, La Chaux-de-Fonds, Switzerland

• Igor A. Guchev, MD, PhD, Respiratory and Internal Disease Department, Branch #4 of the “1586 Regional Military Hospital of The Western Military District” MoD, Smolensk, Russia

• Felix Gutierrez, MD, PhD, Infectious Diseases Unit, Hospital Universitario de Elche, Clinical Medicine Department, Universidad Miguel Hernández, Elche, Alicante, Spain

• Susanne S. Hernes, MD, Department of Geriatrics and Internal Medicine, Sorlandet Hospital Arendal HF, Arendal, Norway

• Andy I.M. Hoepelman, Department of Internal Medicine & Infectious Diseases, University Medical Center, Utrecht, The Netherlands

• Ulla Hohenthal, MD, PhD, Department of Medicine, Turku University Hospital, Turku, Finland.

• Niclas Johansson, MD, Karolinska Institutet, Department of Medicine, Infectious Diseases Unit, Karolinska University Hospital, Stockholm, Sweden.

• Vitezslav Kolek, MD, Prof. PhD, Department of Respiratory Medicine, University Hospital, Palacky University, Olomouc, Czech Republic

• Roman S. Kozlov, MD, MSc, PhD, DSc, Institute of Antimicrobial Chemotherapy (IAC), Smolensk State Medical Academy (SSMA), Smolensk, Russia

• Tsai-Ling Lauderdale, PhD., National Institute of Infectious Diseases and Vaccinology, National Health Research Institutes, Zhunan, Taiwan

• Ivana Mareković, MD, PHD, Department of Clinical and Molecular Microbiology, University Hospital Centre Zagreb, Zagreb, Croatia.

• Mar Masiá, MD, PhD, Infectious Diseases Unit, Hospital Universitario de Elche, Elche, Alicante, Spain

• Matta A. Matta, MD, Infectious Disease Department, Bellevue Medical Center, Mansourieh, Lebanon.

• Òscar Miró, MD, PhD, Emergency Department, Hospital Clínic, and "Emergencies: processes and diseases" Research Group, IDIBAPS; Barcelona, Catalonia, Spain

• David R Murdoch MD, MSc, DTM&H, FRACP, FRCPA, Department of Pathology, University of Otago, Christchurch, New Zealand and Microbiology Unit, Canterbury Health Laboratories, Christchurch, New Zealand

• Eric Nuermberger, MD, Johns Hopkins Medical Institutions, Baltimore, USA

• Richard Paolini, MBBS, MMed (Clin Epi), Department of Emergency Medicine, Concord Repatriation General Hospital, Sydney, Australia

• Rafael Perelló, MD, Emergency Department, Hospital Clinic, C/villarroel 136 08036 Barcelona, Catalonia, Spain.

• Dominic Snijders, MD, Department of Pulmonary Diseases, Alkmaar, The Netherlands

• Vanda Plečko, MD, PHD, Department of Clinical and Molecular Microbiology, University Hospital Centre Zagreb, Zagreb, Croatia.

• Roger Sordé, MD, Department of Infectious Diseases, Hospital Universitari Vall d'Hebron, Vall d'Hebron Research Institute (VHIR), Universitat Autònoma de Barcelona, Spanish Network for Research in Infectious Diseases (REIPI), Barcelona, Spain

• Kristoffer Strålin, MD, PhD, Department of Infectious Diseases, Örebro University Hospital, Örebro, Sweden

• Menno M. van der Eerden, MD, PhD, Department of Pulmonary Diseases, Erasmus MC, Rotterdam The Netherlands

• Angel Vila-Corcoles, MD, Public Health Department, Institut Català de la Salut, Tarragona, Spain

• James P. Watt, MD, MPH, Department of International Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, USA

We gratefully acknowledge helpful advice on the study design and conduct from John Bartlett, Matthew Moore, Christine Segonds, Aimee Summers, and members of the AGEDD Expert Committee (Claire Broome, Bradford D. Gessner, Lee Harrison, Keith Klugman, Diana Lennon, Nancy Messonnier, Andrew J. Pollard, Art Reingold, Igor Rudan, Caroline Trotter, Cynthia Whitney)

Author Contributions

Conceived and designed the experiments: MAS HLJ MD-K KLO BASN. Analyzed the data: MAS HLJ MD-K KLO BASN Group co-authors. Contributed reagents/materials/analysis tools: MAS HLJ MD-K KLO BASN Group co-authors. Wrote the paper: MAS HLJ MD-K KLO BASN Group co-authors.

References

  1. 1. World Health Organization (2007) Pneumococcal conjugate vaccine for childhood immunization--WHO position paper. Wkly Epidemiol Rec 82: 93–104.
  2. 2. O'Brien KL, Wolfson LJ, Watt JP, Henkle E, Deloria-Knoll M, et al. (2009) Burden of disease caused by Streptococcus pneumoniae in children younger than 5 years: global estimates. Lancet 374: 893–902.
  3. 3. Marshall BM, Levy SB (2011) Food animals and antimicrobials: impacts on human health. Clin Microbiol Rev 24: 718–733.
  4. 4. Heffron R (1939) Pneumonia, with special reference to pneumococcus lobar pneumonia. New York: Commonwealth Fund.
  5. 5. Lim WS, Baudouin SV, George RC, Hill AT, Jamieson C, et al. (2009) BTS guidelines for the management of community acquired pneumonia in adults: update 2009. Thorax 64 Suppl 3iii1–55.
  6. 6. Bartlett JG, Breiman RF, Mandell LA, File Jr TM (1998) Community-acquired pneumonia in adults: Guidelines for management. Clinical Infectious Diseases 26: 811–838.
  7. 7. File TM (2003) Community-acquired pneumonia. Lancet 362: 1991–2001.
  8. 8. Mandell LA, Wunderink RG, Anzueto A, Bartlett JG, Campbell GD, et al. (2007) Infectious Diseases Society of America/American Thoracic Society Consensus Guidelines on the management of community-acquired pneumonia in adults. Clinical Infectious Diseases 44: S27–S72.
  9. 9. Bartlett JG (2004) Decline in microbial studies for patients with pulmonary infections. Clinical Infectious Diseases 39: 170–172.
  10. 10. Ruiz-Gonzalez A, Falguera M, Nogues A, Rubio-Caballero M (1999) Is Streptococcus pneumoniae the leading cause of pneumonia of unknown etiology? A microbiologic study of lung aspirates in consecutive patients with community-acquired pneumonia. Am J Med 106: 385–390.
  11. 11. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, et al. (2000) Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. Jama 283: 2008–2012.
  12. 12. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, et al. (2009) The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS medicine 6: e1000100.
  13. 13. Fine MJ, Auble TE, Yealy DM, Hanusa BH, Weissfeld LA, et al. (1997) A prediction rule to identify low-risk patients with community-acquired pneumonia. N Engl J Med 336: 243–250.
  14. 14. DerSimonian R, Laird N (1986) Meta-analysis in Clinical Trials. Controlled Clinical Trials 7: 177–188.
  15. 15. Higgins JP, Thompson SG (2004) Controlling the risk of spurious findings from meta-regression. Statistics in medicine 23: 1663–1682.
  16. 16. StataCorp (2009) Stata Statistical Software: Release 11. College Station, TX: StataCorp LP
  17. 17. Efron B, Tibshirani RJ (1993) An Introduction to the Bootstrap. Boca Raton: Chapman & Hall/CRC.
  18. 18. Austrian R, Gold J (1964) Pneumococcal Bacteremia with Especial Reference to Bacteremic Pneumococcal Pneumonia. Ann Intern Med 60: 759–776.
  19. 19. Mufson MA (1981) Pneumococcal infections. Jama 246: 1942–1948.
  20. 20. Tilghman RC, Finland M (1937) Clinical significance of bacteremia in pneumococcic pneumonia. Archives of Internal Medicine 59: 602–619.
  21. 21. Murdoch DR, Laing RT, Mills GD, Karalus NC, Town GI, et al. (2001) Evaluation of a rapid immunochromatographic test for detection of Streptococcus pneumoniae antigen in urine samples from adults with community-acquired pneumonia. J Clin Microbiol 39: 3495–3498.
  22. 22. Fine MJ, Smith MA, Carson CA, Mutha SS, Sankey SS, et al. (1996) Prognosis and outcomes of patients with community-acquired pneumonia. A meta-analysis. Jama 275: 134–141.
  23. 23. Simonsen L, Taylor RJ, Young-Xu Y, Haber M, May L, et al. (2011) Impact of pneumococcal conjugate vaccination of infants on pneumonia and influenza hospitalization and mortality in all age groups in the United States. MBio 2.
  24. 24. Jokinen J, Scott JA (2010) Estimating the proportion of pneumonia attributable to pneumococcus in Kenyan adults: latent class analysis. Epidemiology 21: 719–725.
  25. 25. Albrich WC, Muller B, Harbarth S (2010) Antibiotic therapy in patients hospitalized with acute chronic obstructive pulmonary disease. JAMA: the journal of the American Medical Association 304: 1326; author reply 1326–1327.
  26. 26. Waterer GW, Wunderink RG (2001) The influence of the severity of community-acquired pneumonia on the usefulness of blood cultures. Respir Med 95: 78–82.
  27. 27. Feldman C, Glatthaar M, Morar R, Mahomed AG, Kaka S, et al. (1999) Bacteremic pneumococcal pneumonia in HIV-seropositive and HIV-seronegative adults. Chest 116: 107–114.
  28. 28. Hirschtick RE, Glassroth J, Jordan MC, Wilcosky TC, Wallace JM, et al. (1995) Bacterial pneumonia in persons infected with the human immunodeficiency virus. Pulmonary Complications of HIV Infection Study Group. N Engl J Med 333: 845–851.
  29. 29. Roson B (2004) Contribution of a urinary antigen assay (Binax NOW) to the early diagnosis of pneumococcal pneumonia. Clin Infect Dis 38: 222–226.
  30. 30. Gutiérrez F, Masiá M, Rodríguez JC, Ayelo A, Soldán B, et al. (2003) Evaluation of the immunochromatographic Binax NOW for detection of Streptococcus pneumoniae urinary antigen in a prospective study of community-acquired pneumonia in Spain. Clinical Infectious Diseases 36: 286–292.
  31. 31. Stralin K, Kaltoft MS, Konradsen HB, Olcen P, Holmberg H (2004) Comparison of two urinary antigen tests for establishment of pneumococcal etiology of adult community-acquired pneumonia. J Clin Microbiol 42: 3620–3625.
  32. 32. Ortega L, Sierra M, Dominguez J, Martinez J, Matas L, et al. (2005) Utility of a pneumonia severity index in the optimization of the diagnostic and therapeutic effort for community-acquired pneumonia. Scand J Infect Dis 37: 657–663.
  33. 33. Levy M, Dromer F, Brion N, Leturdu F, Carbon C (1988) Community-acquired pneumonia. Importance of initial noninvasive bacteriologic and radiographic investigations. Chest 93: 43–48.
  34. 34. Domínguez J, Galí N, Blanco S, Pedroso P, Prat C, et al. (2001) Detection of Streptococcus pneumoniae antigen by a rapid immunochromatographic assay in urine samples. Chest 119: 243–249.
  35. 35. Gutierrez F, Masía M, Rodriguez JC, Ayelo A, Soldan B, et al. (2003) Evaluation of the immunochromatographic binax NOW assay for detection of Streptococcus pneumoniae urinary antigen in a prospective study of community-acquired pneumonia in Spain. Clinical Infectious Diseases 36: 286–292.
  36. 36. Smith MD, Derrington P, Evans R, Creek M, Morris R, et al. (2003) Rapid diagnosis of bacteremic pneumococcal infections in adults by using the Binax NOW Streptococcus pneumoniae urinary antigen test: a prospective, controlled clinical evaluation. J Clin Microbiol 41: 2810–2813.
  37. 37. Pilishvili T, Lexau C, Farley MM, Hadler J, Harrison LH, et al. (2010) Sustained reductions in invasive pneumococcal disease in the era of conjugate vaccine. J Infect Dis 201: 32–41.
  38. 38. Grijalva CG, Nuorti JP, Arbogast PG, Martin SW, Edwards KM, et al. (2007) Decline in pneumonia admissions after routine childhood immunisation with pneumococcal conjugate vaccine in the USA: a time-series analysis. Lancet 369: 1179–1186.
  39. 39. Marekovic I, Plecko V, Boras Z, Pavlovic L, Budimir A, et al. (2008) Evaluation of PCR as rapid microbiological method in diagnosis of pneumococcal pneumonia. Scandinavian Journal of Infectious Diseases 40: 843–845.
  40. 40. Krcova M, Fojtu H, Fila L, Hrazdirova A, Hutyrova B, et al. New approach to detecting Streptococcus pneumoniae in community-acquired pneumonia. Studia Pneumologica et Phthiseologica 2008 68: 4: 151–155.
  41. 41. Lim WS, Macfarlane JT, Boswell TC, Harrison TG, Rose D, et al. (2001) Study of community acquired pneumonia aetiology (SCAPA) in adults admitted to hospital: implications for management guidelines. Thorax 56: 296–301.
  42. 42. Hohenthal U, Vainionpaa R, Meurman O, Vahtera A, Katiskalahti T, et al. (2008) Aetiological diagnosis of community acquired pneumonia: utility of rapid microbiological methods with respect to disease severity. Scand J Infect Dis 40: 131–138.
  43. 43. Lasocki S, Scanvic A, Le Turdu F, Restoux A, Mentec H, et al. (2006) Evaluation of the Binax NOW Streptococcus pneumoniae urinary antigen assay in intensive care patients hospitalized for pneumonia. Intensive Care Med 32: 1766–1772.
  44. 44. Endeman H, Schelfhout V, Voorn GP, van Velzen-Blad H, Grutters JC, et al. (2008) Clinical features predicting failure of pathogen identification in patients with community acquired pneumonia. Scand J Infect Dis 40: 715–720.
  45. 45. Snijders D, Daniels JM, de Graaff CS, van der Werf TS, Boersma WG (2010) Efficacy of corticosteroids in community-acquired pneumonia: a randomized double-blinded clinical trial. Am J Respir Crit Care Med 181: 975–982.
  46. 46. van der Eerden MM, Vlaspolder F, de Graaff CS, Groot T, Jansen HM, et al. (2005) Value of intensive diagnostic microbiological investigation in low- and high-risk patients with community-acquired pneumonia. Eur J Clin Microbiol Infect Dis 24: 241–249.
  47. 47. Hernes SS, Hagen E, Tofteland S, Finsen NT, Christensen A, et al. (2010) Transthoracic fine-needle aspiration in the aetiological diagnosis of community-acquired pneumonia. Clin Microbiol Infect 16: 909–911.
  48. 48. Guchev IA, Yu VL, Sinopalnikov A, Klochkov OI, Kozlov RS, et al. (2005) Management of nonsevere pneumonia in military trainees with the urinary antigen test for Streptococcus pneumoniae: an innovative approach to targeted therapy. Clin Infect Dis 40: 1608–1616.
  49. 49. Beovic B, Bonac B, Kese D, Avsic-Zupanc T, Kreft S, et al. (2003) Aetiology and clinical presentation of mild community-acquired bacterial pneumonia. Eur J Clin Microbiol Infect Dis 22: 584–591.
  50. 50. Andreo F, Dominguez J, Ruiz J, Blanco S, Arellano E, et al. (2006) Impact of rapid urine antigen tests to determine the etiology of community-acquired pneumonia in adults. Respir Med 100: 884–891.
  51. 51. Falguera M, Trujillano J, Caro S, Menendez R, Carratala J, et al. (2009) A prediction rule for estimating the risk of bacteremia in patients with community-acquired pneumonia. Clin Infect Dis 49: 409–416.
  52. 52. Falguera M, Ruiz-Gonzalez A, Schoenenberger JA, Touzon C, Gazquez I, et al. (2010) Prospective, randomised study to compare empirical treatment versus targeted treatment on the basis of the urine antigen results in hospitalised patients with community-acquired pneumonia. Thorax 65: 101–106.
  53. 53. Fernandez-Sabe N, Carratala J, Roson B, Dorca J, Verdaguer R, et al. (2003) Community-acquired pneumonia in very elderly patients: causative organisms, clinical characteristics, and outcomes. Medicine (Baltimore) 82: 159–169.
  54. 54. Marcos MA, de Anta MTJ, de la Bellacasa JP, Gonzalez J, Martinez E, et al. (2003) Rapid urinary antigen test for diagnosis of pneumococcal community-acquired pneumonia in adults. European Respiratory Journal 21: 209–214.
  55. 55. Perello R, Miro O, Miro JM, Masso E, Sanchez M, et al. (2010) Urgent pneumococcal urinary antigen, importance in the diagnosis of acquired pneumonia in HIV-1 patients. Eur J Emerg Med
  56. 56. Sorde R, Falco V, Lowak M, Domingo E, Ferrer A, et al. (2011) Current and potential usefulness of pneumococcal urinary antigen detection in hospitalized patients with community-acquired pneumonia to guide antimicrobial therapy. Arch Intern Med 171: 166–172.
  57. 57. Vidal J, Adrover A, Garcia-Mendez E, Garcia-Rey C, Dal-Re R (2006) Lack of usefulness of the immunochromatographic urinary pneumococcal antigen test in maximizing pneumococcal yield in a community-acquired pneumonia clinical trial. Journal of Chemotherapy 18: 562–564.
  58. 58. Johansson N, Kalin M, Annika TL, Giske CG, Hedlund J (2010) Etiology of Community-Acquired pneumonia: Increased microbiological yield with new diagnostic methods. Clinical Infectious Diseases 50: 202–209.
  59. 59. Ehara N, Fukushima K, Kakeya H, Mukae H, Akamatsu S, et al. (2008) A novel method for rapid detection of Streptococcus pneumoniae antigen in sputum and its application in adult respiratory tract infections. Journal of Medical Microbiology 57: 820–826.
  60. 60. Ishida T, Hashimoto T, Arita M, Ito I, Osawa M (1998) Etiology of community-acquired pneumonia in hospitalized patients: a 3-year prospective study in Japan. Chest 114: 1588–1593.
  61. 61. Kobashi Y, Yoshida K, Miyashita N, Niki Y, Matsushima T (2007) Evaluating the use of a Streptococcus pneumoniae urinary antigen detection kit for the management of community-acquired pneumonia in Japan. Respiration 74: 387–393.
  62. 62. Lauderdale TL, Chang FY, Ben RJ, Yin HC, Ni YH, et al. (2005) Etiology of community acquired pneumonia among adult patients requiring hospitalization in Taiwan. Respir Med 99: 1079–1086.
  63. 63. Charles PG, Whitby M, Fuller AJ, Stirling R, Wright AA, et al. (2008) The etiology of community-acquired pneumonia in Australia: why penicillin plus doxycycline or a macrolide is the most appropriate therapy. Clin Infect Dis 46: 1513–1521.
  64. 64. Weatherall C, Paoloni R, Gottlieb T (2008) Point-of-care urinary pneumococcal antigen test in the emergency department for community acquired pneumonia. Emerg Med J 25: 144–148.
  65. 65. Diaz A, Barria P, Niederman M, Restrepo MI, Dreyse J, et al. (2007) Etiology of community-acquired pneumonia in hospitalized patients in chile: the increasing prevalence of respiratory viruses among classic pathogens. Chest 131: 779–787.
  66. 66. Matute AJ, Brouwer WP, Hak E, Delgado E, Alonso E, et al. (2006) Aetiology and resistance patterns of community-acquired pneumonia in Leon, Nicaragua. Int J Antimicrob Agents 28: 423–427.
  67. 67. Butler JC, Bosshardt SC, Phelan M, Moroney SM, Tondella ML, et al. (2003) Classical and latent class analysis evaluation of sputum polymerase chain reaction and urine antigen testing for diagnosis of pneumococcal pneumonia in adults. J Infect Dis 187: 1416–1423.
  68. 68. Nuermberger E, Khalil A, Rothman R, Auwaerter P (2005) Performance of a Streptococcus pneumoniae Urinary Antigen Test (UAT) in HIV-Infected Patients with Community-Acquired Pneumonia (CAP). American Thoracic Society International Conference. San Diego
  69. 69. Watt JP, MoiSi JC, Donaldson RLA, Reid R, Ferro S, et al. (2010) Use of serology and urine antigen detection to estimate the proportion of adult community-acquired pneumonia attributable to Streptococcus pneumoniae. Epidemiology and Infection 1–8.