|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Departamento de Enfermedades Respiratorias (Drs. Díaz, Barria, and Saldias, Mr. Dreyse, Mr. Fuentes, and Ms. Couble), Pontificia Universidad Católica de Chile, Santiago, Chile; the Department of Medicine (Dr. Niederman), Winthrop University Hospital, Mineola, NY; and the Department of Medicine (Dr. Restrepo), Division of Pulmonary/Critical Care Medicine, and Infectious Diseases, South Texas Veterans Health Care System, San Antonio, TX.
Correspondence to: Alejandro Díaz, MD, 20 Bonner Ave, Somerville, MA 02143; e-mail: alediazf{at}hotmail.com
Abstract
Background and study objectives: The range and relative impact of microbial pathogens, particularly viral pathogens, as a cause of community-acquired pneumonia (CAP) in hospitalized adults has not received much attention. The aim of this study was to determine the microbial etiology of CAP in adults and to identify the risk factors for various specific pathogens.
Methods: We prospectively studied 176 patients (mean [± SD] age, 65.8 ± 18.5 years) who had hospitalized for CAP to identify the microbial etiology. For each patient, sputum and blood cultures were obtained as well as serology testing for Mycoplasma pneumoniae and Chlamydophila pneumoniae, urinary antigen testing for Legionella pneumophila and Streptococcus pneumoniae, and a nasopharyngeal swab for seven respiratory viruses.
Results: Microbial etiology was determined in 98 patients (55%). S pneumoniae (49 of 98 patients; 50%) and respiratory viruses (32%) were the most frequently isolated pathogen groups. Pneumococcal pneumonia was associated with tobacco smoking of > 10 pack-years (odds ratio [OR], 2.6; 95% confidence interval [CI], 1.2 to 5.4; p = 0.01). Respiratory viruses were isolated more often in fall or winter (28%; p = 0.011), and as an exclusive etiology tended to be isolated in patients
65 years of age (20%; p = 0.07). Viral CAP was associated with antimicrobial therapy prior to hospital admission (OR, 4.5; 95% CI, 1.4 to 14.6).
Conclusions: S pneumoniae remains the most frequent pathogen in adults with CAP and should be covered with empirical antimicrobial treatment. Viruses were the second most common etiologic agent and should be tested for, especially in fall or winter, both in young and elderly patients who are hospitalized with CAP.
Key Words: bacterial pneumonia community etiology viruses
Community-acquired pneumonia (CAP) is the leading cause of death from infectious diseases both in developed1 and in developing countries.2 According to authoritative guidelines for the management of patients with CAP guidelines,345 detailed information on its etiology is required for the formulation of treatment recommendations and preventive measures.
Despite vigorous clinical investigation, the etiology of CAP remains unknown in 30 to 60% of cases.6 Most studies show that Streptococcus pneumoniae remains the primary cause of CAP.7891011121314151617 In the last 2 decades, the following several factors have potentially affected microbial patterns of CAP adult in Chile: (1) increased aging of the population, often with increased prevalence of comorbid illnesses, has increased concerns about Gram-negative bacilli infection in CAP18; (2) the emergence of "new" pathogens such as Chlamydophila pneumoniae19; and (3) the increasing use of antibiotics such as macrolides20 and the resulting increased antibiotic resistance. Moreover, respiratory viruses have received more recognition as pathogens in adults with CAP.2122 The impact of these changes on the microbial etiology of CAP in Chile and how they should influence new management guidelines is unknown. The aim of the present study was to determine etiology of CAP, and the risk factors associated with specific pathogens.
Materials and Methods
A prospective cohort study of consecutive patients admitted to the hospital with CAP was conducted between February 27, 2003, and April 15, 2005, at a 520-bed university-affiliated teaching hospital in Chile. The study received institutional review board approval, and informed consent was obtained from all patients. Subjects were included if they were not immunocompromised, were
16 years of age, and had radiographically confirmed pneumonia requiring hospitalization. CAP was defined as the presence of acute respiratory symptoms, fever, or altered mental status, a new infiltrate found on chest radiograph, and no emerging alternative diagnosis during the follow-up. Exclusion criteria included the following: transfer from another hospital; hospitalization in the preceding 30 days; immunosuppressive illness (ie, active treatment for cancer, HIV positive, use of > 20 mg/d prednisone or other immunosuppressive agents, and history of organ transplant); and active tuberculosis.
Data Collection at Study Entry
Data collected at hospital admission included place of residency, antibiotic treatment prior to admission, age, gender, comorbid illnesses, and smoking status. Clinical symptoms, vital signs, site of hospital admission, and antimicrobial treatment were also recorded. Patient severity of illness was assessed by the pneumonia severity index (PSI) as described by Fine at al.23 Treatment was considered adequate if at least one administered antibiotic was active against the isolated pathogen. Chest radiographic findings were classified as alveolar, interstitial, or mixed infiltrate. Extension (ie, the number of lobes involved and bilaterality), cavitation, and the presence of pleural effusion were also recorded. Recorded complications included the following: use of mechanical ventilation; septic shock24; and empyema (defined as infected pleural fluid that was drained by a chest tube). The 30-day mortality rate was also recorded.
Microbiological Evaluation
The following tests were obtained on hospital admission and prior to antibiotic administration: two sets of blood cultures (bioMerieux Inc; Durham, NC); and sputum for culture. A representative lower respiratory tract sample was defined as containing > 25 leukocytes and < 10 epithelial cells per low-power field. All bacterial species isolated were identified by standard techniques. In the first few days after hospital admission and 4 to 6 weeks thereafter, sera were collected, and tested for the presence of IgG and IgM to Mycoplasma pneumoniae (Zeus Scientific Inc; Raritan, NJ), C pneumoniae, and Chlamydophila psittaci (Vircell Inc; Santa Fe, Granada). Antibodies to M pneumoniae and Chlamydophila spp were tested for with indirect immunofluorescence and indirect microimmunofluorescence, respectively. Nasopharyngeal swabs for the detection of seven viral antigens were obtained, transported, and processed on the day of collection. Tests to detect influenza virus types A and B, parainfluenza virus types 1 to 3, adenovirus, and respiratory syncytial virus were performed with monoclonal antibodies for direct immunofluorescence assay (DFA) (Remel Inc; Lenexa, KA). The collected urine samples were stored at 70°C or 20°C until testing. The detection of Legionella pneumophila by means of an enzyme immunoassay system and S pneumoniae by means of an immunochromatographic membrane assay (Binax Inc; Scarborough, ME) in a nonconcentrated urine sample was performed.
Diagnostic Criteria
An etiologic diagnosis was considered to be definitive if one of the following criteria were met: (1) isolation of respiratory pathogen in a sterile specimen (blood and pleural fluid)1013; (2): fourfold rise in IgG titers for M pneumoniae, C pneumoniae (
1:512), and C psittaci (
1:64)13; (3) a single increased IgM titer for M pneumoniae (
1:16) or C pneumoniae (
1:10) [according to the manufacturer instructions]; (4) positive urinary antigen for L pneumophila type 113 or S pneumoniae25; and (5) a positive result for one respiratory virus. An etiologic diagnosis was considered to be presumptive if a validated sputum sample yielded one or more bacterial strains,13 or a single increased IgG titer for C pneumoniae (
1:512).
Statistical Analysis
Data were analyzed using a statistical software package (Minitab; State College, PA). Results are expressed as the mean ± SD. Univariate analysis was carried out using the
2 test and Fisher exact test for categoric data, and the t test for independent samples for continuous variables. Pathogens were grouped in etiologic categories as follows: S pneumoniae; viruses; atypical microbial agents (ie, M pneumoniae, C pneumoniae, and L pneumophila); mixed infections; and undetermined etiology. Univariate associations among these categories and patient age, smoking habits, antibiotic use prior to hospital admission, and comorbid illnesses were tested. Variables significantly associated with each etiologic category on univariate analysis were incorporated into a stepwise multiple logistic regression model to identify independent risk factors for each etiologic category. The results are expressed as odds ratios (ORs) with 95% confidence intervals (CIs). All reported p values are two-tailed, and the level of significance was set at 5%.
Results
Patient Characteristics
One hundred seventy-six patients (91 men and 85 women; mean age, 65.8 ± 18.5 years; range, 17 to 101 years) were studied. The principal clinical characteristics and chest radiograph findings are summarized in Table 1
. The mean duration of clinical symptoms before hospital admission was 7.1 ± 5.7 days. Thirty percent of patients had received an antibiotic prior to hospital admission, and 20% were admitted to the ICU.
|
|
S pneumoniae was identified exclusively by sputum testing in 5 patients (10%), by blood cultures in 12 patients (24%), by urine sample in 25 patients (51%), and by more than one technique in 7 patients (14%). Among 150 viral panels collected, the results of 32 panels (21%) were positive (with another bacterial pathogen found in 9 patients).
Microbial Etiology of CAP Patients
The etiology was identified in 98 patients (55%) [13 patients (13%) with two or more pathogens], and identification was more likely in patients > 45 years of age (81%) than in younger patients (19%; p = 0.031). The top five most frequent pathogens (Table 3
) isolated as a single agent were S pneumoniae (43 patients; 44%), parainfluenza virus types 1 to 3 (13 patients; 13%), influenza virus types A and B (8 patients; 8%); Haemophilus influenzae (5 patients; 5%), and L pneumophila (4 patients; 4%).
|
Respiratory viruses were found in 32 patients (32%) [exclusively in 23 patients]. Viruses were isolated more often during the fall or winter than during other seasons (28% vs 6%, respectively; p = 0.011) [Fig 1
] and tended to be isolated in patients
65 years of age (17 of 84 patients; 20%; p = 0.07).
|
|
65 years) or specific comorbid conditions, including diabetes mellitus, and respiratory, cardiac, neurologic, hepatic, and renal diseases. Tobacco smoking of > 10 pack-years was significantly associated with pneumococcal pneumonia (OR, 2.6; 95% CI, 1.2 to 5.4; p = 0.01), and antimicrobial treatment within a month prior to hospital admission was associated with viral pneumonia (OR, 4.5; 95% CI, 1.4 to 14.6; p = 0.01).
Antimicrobial Therapy and Hospital Course
Seventy patients (40%) received empirical monotherapy with cephalosporins (83%), respiratory fluoroquinolones (levofloxacin or moxifloxacin) [14%], or macrolides (3%), and 105 patients (60%) received a third-generation cephalosporin plus a respiratory fluoroquinolone (49%), a macrolide (39%), or an antianaerobic (8%) antibiotic. One patient received three antibiotics. Sixty patients (34%) were receiving therapy with empiric antibiotics administered according to American Thoracic Society guidelines.3 The 30-day mortality rate among the 71 patients who received adequate antimicrobial treatment was lower than that among those who did not (8.4% vs 25%, respectively; p = 0.33). The results of main outcomes in patients with determined etiology compared to those with undetermined etiology were as follows: one or more complications, 17% vs 13%, respectively; mean of length of stay, 9.4 ± 5.1 vs 7.4 ± 5.9 days, respectively [p = 0.017]; and 30-day mortality rate, 9% vs 13%, respectively. Outcomes are summarized in Table 5
.
|
This study is the first to document the relative importance of various pathogens in causing CAP in hospitalized patients in Chile, a developing country. The main results were as follows: S pneumoniae was the most frequent pathogen found followed by respiratory viruses; viruses were more common in elderly patients, and were isolated mainly in the fall or winter; and viral CAP patients had significant mortality.
The etiology was found in 55% of cases, and this figure is similar to those from some other studies,811131415 but is lower than those from others.7101216 Several reasons may explain this result, as follows: a high proportion of patients who had received prior antimicrobial treatment and a low number of paired serology samples collected in this study could have reduced the diagnostic yield in our cohort; some techniques such as polymerase chain reaction for atypical microbial agents or viruses were not used; and pathogens such as other Legionella species and Coxiella burnetti were not included. Studies121622 that included these techniques or pathogens identified an etiologic agent in 75 to 80.6% of cases.
S pneumoniae is the most frequent microorganism isolated in adult patients as has been the case in most studies7891012131415161718 in hospitalized patients with CAP. One of these studies17 performed in a neighboring country to Chile also showed S pneumoniae as the leading microorganism. This similarity suggests that this CAP etiology feature may be common in several countries throughout South America. Our finding also confirms that the current local guidelines for the empirical treatment of CAP must cover this organism in inpatients. We found that tobacco smoking of > 10 pack-years was associated with pneumococcal pneumonia. This finding is consistent with those of prior studies that have demonstrated that tobacco smoking is a risk factor for invasive pneumococcal disease26 and bacteremic pneumococcal pneumonia.27 Changes in mucociliary clearance, bacterial adherence,28 and respiratory epithelium29 have been suggested mechanisms by which exposure to tobacco smoke increases the risk of pneumococcal disease. The current finding emphasizes that smoking cessation should be recommended by physicians who treat CAP patients.
We found that 89% of cases of bacteremic pneumonia were caused by S pneumoniae, and this figure confirms data from a previous metaanalysis,30 which reported that this pathogen is the main cause of bacteremic pneumonia. However, the mortality rate of patients with pneumococcal pneumonia in the current study was lower than that reported in previous studies.30313233 This difference may be explained by the fact that patients in our study were younger than those in previous studies, and a lower proportion either had a high risk of disease or was admitted to the ICU. Interestingly, S pneumoniae was found more often in blood cultures than in sputum cultures. This finding may support the usefulness of blood cultures as an easy and inexpensive etiologic diagnostic technique in hospitalized patients with CAP.
The rate of S pneumoniae strain resistance to ß-lactamic antimicrobial agents was lower than previously reported rates,3132 and the current low resistance rate (4%) to cefotaxime supports its use as empirical treatment for CAP requiring hospitalization in Chile; the rate of erythromycin-resistant S pneumoniae was 25% by contrast. A partial explanation for this finding may be the increased prescription of macrolides during previous years in Chile.20
One of the most instructive findings in the current study was the high proportion of respiratory viruses isolated in CAP patients in whom an etiologic diagnosis was made (32%; 23% as a unique microbial agent). Our finding provides information that demonstrates a possible pathogenic role for viruses in adults with CAP and supports the findings of previous CAP studies.7891011121314161722 The main characteristics of patients with a viral agent were age
65 years, hospital admission during the fall or winter, presence of comorbid illnesses, and a mortality rate of approximately 9%. These results are in contrast to those from a previous report34 that observed viruses had no significant seasonal predominance, in which no patient died.
We considered a positive DFA result as diagnostic of a viral etiology based on prior literature7891011121314161722 that reported viral pathogens causing CAP in adults, and a study of > 1,500 respiratory samples that reported that DFA was equivalent to cell culture in the detection of respiratory viruses.35 Therefore, our finding provides evidence to consider using DFA for the detection of multiple common respiratory viruses in CAP patients. Compared to viral multiplex polymerase chain reaction assay,22 DFA is less expensive and easier to process, but it has less sensitivity.36 Compared to the findings of viral serology and viral cultures, DFA allows a clinically relevant shorter time to obtain results.
We found that the administration of antimicrobial treatment prior to hospital admission was a risk factor for viral pneumonia. A reason for this might be that those patients had a mixed bacterial and viral infection, and antibiotics eliminated bacteria, giving the false impression that there was only a viral infection. Other findings such as the frequency of symptoms, PSI score, and mortality rate of patients with viral pneumonia, which were similar to other etiologic diagnostic data, support this possibility. The practical implications of our data on viral pneumonia could be that testing for these pathogens should be sought in fall or winter and not only in young patients, but also in elderly patients with CAP requiring hospitalization.
Atypical bacterial pathogens, as a single infection, were the third most frequent cause (10%) of CAP in our patient cohort. The occurrences of both L pneumophila and C pneumoniae in this study were less frequent than those in previous studies.1937 Although yearly variation may partly explain this tendency for Legionella infection, an increased use of macrolides in Chile might also be considered.20 Most patients with atypical microbial agents had a high-risk severity score and were older (
60 years of age); half of them had been admitted to intermediate care units or ICUs. These findings provide support for a confirmation that atypical microbial agents should be covered in patients with severe CAP, as has been recommended by CAP guidelines.345 One limitation of this study should be emphasized. Since the number of cases for some etiologies (ie, viruses and atypical pathogens) was small, the incidence of those etiologies should be considered carefully.
In conclusion, the main pathogens identified in CAP patients were S pneumoniae and viruses. Therefore, empirical antibiotic therapy should be effective against S pneumoniae. During the fall or winter, respiratory viruses should be considered as etiologic agents of CAP requiring hospitalization in both young and elderly patients.
Footnotes
Abbreviations: CAP = community-acquired pneumonia; CI = confidence interval; DFA = direct immunofluorescence assay; OR = odds ratio; PSI = pneumonia severity index
This research was supported by a grant from Dirección de Investigación de la P. Universidad Católica de Chile (DIPUC 2003/10E) and by investigation grants from Sociedad Chilena de Enfermedades Respiratorias (2002 and 2004).
The authors have reported to the ACCP that no significant conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.
Received for publication September 19, 2006. Accepted for publication September 28, 2006.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |