Skip Navigation

Etiology, Reasons for Hospitalization, Risk Classes, and Outcomes of Community-Acquired Pneumonia in Patients Hospitalized on the Basis of Conventional Admission Criteria

  1. Beatriz Rosón1,
  2. Jordi Carratalà1,
  3. Jordi Dorca2,
  4. Aurora Casanova3,
  5. Frederic Manresa2, and
  6. Francesc Gudiol1
  1. 1 Infectious Disease, Ciutat Sanitària i Universitària de Bellvitge, University of Barcelona, L'Hospitalet de Llobregat, Barcelona, Spain
  2. 2 Respiratory, Ciutat Sanitària i Universitària de Bellvitge, University of Barcelona, L'Hospitalet de Llobregat, Barcelona, Spain
  3. 3 Microbiology Services, Ciutat Sanitària i Universitària de Bellvitge, University of Barcelona, L'Hospitalet de Llobregat, Barcelona, Spain
  1. Reprints or correspondence: Dr. Francesc Gudiol, Infectious Disease Service, Hospital de Bellvitge, Feixa Llarga s/n, 08907 L'Hospitalet, Barcelona, Spain (fgudiol{at}csub.scs.es).
 
Next Section

Abstract

We performed an observational analysis of prospectively collected data on 533 nonseverely immunosuppressed adult patients who were hospitalized for community-acquired pneumonia on the basis of conventional admission criteria. For this population, we correlated etiology, reasons for admission, and outcomes using the Pneumonia Severity Index (PSI), to identify major discrepancies between the PSI risk class and the conventional criteria for deciding the site of care. PSI classes and corresponding mortality rates were as follows: class I, 51 patients (0%); class II, 62 (2%); class III, 117 (3%); class IV, 198 (10%); and class V, 105 (29%). We identified significant discrepancies between both methods. Overall, 230 patients (40%) who were hospitalized according to conventional criteria were assigned to low-risk classes. Of these 230 patients, 137 (60%) needed supplementary oxygen or had pleural complications; for the remaining patients, there were no irrefutable reasons for admission. This latter group deserves prospective evaluation in randomized studies that compare ambulatory and in-hospital management.

Management of community-acquired pneumonia (CAP) has been progressively complicated by several factors, such as the expanding spectrum of causative organisms, the rising prevalence of resistance to antimicrobial agents, the increasing population of patients of advanced age and with comorbidities, and the interest in reducing the number of unnecessary hospitalizations [1–3]. From the standpoints of cost and quality of care, a key decision is whether to treat patients as outpatients or as inpatients [4]. At the present time, most hospitals implement protocols based on conventional admission criteria, and physicians make their decisions according to their experience and subjective clinical judgement. However, there is general agreement that a substantial number of patients who are currently hospitalized because of CAP on the basis of conventional criteria might be safely treated as outpatients [5–7]. Because of the variability in determining the need for hospitalization, several prognostic scoring and outcome assessment tools have been developed [8–11].

In 1997, the Pneumonia Outcomes Research Team developed a prediction rule based on a Pneumonia Severity Index (PSI) that stratified patients into 5 risk classes and showed a direct correlation between risk class and death [12]. This prediction rule accurately identifies patients with CAP who are at low risk of death, and it has been proposed as a guide for decisions regarding hospitalization. Nevertheless, this stratification may not be well suited for deciding the site of care for some low-risk-class patients, because a number of factors that may greatly influence the decision to hospitalize are not considered by the rule [13].

Since 1995, we have performed a prospective evaluation of all nonseverely immunosuppressed adult patients with CAP who have been hospitalized on the basis of the conventional admission criteria currently in use in the emergency department of our hospital. The aims of the present study were to correlate the PSI with the reasons for admission, etiology, and outcomes for patients hospitalized on the basis of these criteria, and to identify major discrepancies between these 2 approaches for deciding the site of care.

Previous SectionNext Section

Patients and Methods

Setting and study design. The study was conducted at Bellvitge Hospital, a 1000-bed university hospital for adult patients in Barcelona, which serves an area of 1,100,000 inhabitants. All nonseverely immunosuppressed adult patients with CAP who were admitted to the hospital from February 1995 through May 1997 were prospectively recruited and followed up. All of them fulfilled ⩾1 of the following conventional criteria for hospitalization that are currently used at our institution: age of >70 years; respiratory failure (partial pressure of oxygen, arterial [PaO2], <60 mm Hg or PaO2/fraction of inspired oxygen, <300); multilobar or bilateral radiological involvement; pleural effusion or empyema; hypotension or shock; presence of comorbidities, such as chronic obstructive lung disease, congestive heart failure, renal failure, neoplasia, splenectomy, chronic liver disease, alcoholism, and diabetes mellitus; suspicion of infection with a high-risk pathogen (e.g., Legionella species or gram-negative bacilli) or aspiration pneumonia; septic metastatic lesions; and lack of response to previous antibiotic therapy. Pregnant women, patients with neutropenia or AIDS, and patients who have undergone transplantation were not included in the study. No changes in hospital admission policy were made as the study progressed.

Clinical evaluation and follow-up. At the initial visit and before starting antibiotic therapy, patients underwent a complete clinical history and physical examination. Basic chemistry (for blood urea nitrogen, glucose, sodium, and potassium values) and hematology tests (for hematocrit levels and leukocyte and platelet counts), arterial blood gas determinations, and chest radiography were performed. Two sets of blood samples were obtained and cultured and, when available, a sputum sample was evaluated by use of Gram staining and culture. Invasive procedures, such as bronchoscopy or transthoracic needle aspiration (TNA) were performed if indicated by the attending physician. Paired serum samples obtained during the acute and convalescent phases of infection (separated by 3–8 weeks) were also obtained for serological studies. Patients were seen daily during their hospital stay by one of the investigators (B. R., in most cases), and a long-term follow-up visit took place ∼1 month after discharge.

Microbiological studies. Investigation of pathogens in samples of blood, normally sterile fluids, sputum, and TNA was performed by use of standard microbiological procedures. Latex agglutination for Streptococcus pneumoniae in TNA samples was performed by use of the Slidex Pneumokit (bioMérieux) according to the manufacturer's instructions. Detection of S. pneumoniae in TNA samples was performed by use of PCR, as described elsewhere [14]. Isolation of Legionella pneumophila in samples of sputum and TNA was attempted by use of selective medium (buffered charcoal yeast extract-α). Detection of L. pneumophila serogroup 1 antigen in urine was performed by means of an immunoenzymatic commercial method (Legionella Urinary Antigen, Binax). Standard serological methods were used to detect antibodies to the following pathogens: Mycoplasma pneumoniae (indirect agglutination), Chlamydia psittaci (immunofluorescence [IF]), Chlamydia pneumoniae (micro-IF), Coxiella burnetii (IF), L. pneumophila serogroups 1–6 (EIA), respiratory syncytial virus (EIA), parainfluenza 3 virus (EIA), and influenza A virus (EIA).

Definitions. “CAP” was defined as an acute illness associated with ⩾1 of the following respiratory signs and symptoms: new cough with or without sputum production, pleuritic chest pain, dyspnea, fever or hypothermia, altered breath sounds on auscultation, and the presence of a new infiltrate on a chest radiograph. Acute severity was assessed by the simplified acute physiology score (SAPS), as described elsewhere [15]. “Complications” were defined as any untoward circumstances that occurred during hospitalization, with the exception of the side effects of medication. “Overall mortality” was defined as deaths due to any cause within 30 days of hospitalization.

Classification of pneumonias. Final diagnoses of cases were established by a consensus of clinical investigators after analyzing all the clinical, radiological, and microbiological data and according to the following criteria. Etiologic diagnosis was considered definitive in the following situation: isolation of a respiratory pathogen in a normally sterile specimen, such as blood, pleural fluid, or TNA; isolation of L. pneumophila or Mycobacterium tuberculosis in sputum samples; positive results of latex agglutination tests for pneumococcal antigens or positive pneumococcal DNA determination by use of PCR in pleural fluid or TNA specimens; detection of L. pneumophila serogroup 1 antigen in urine; presence of IgM antibodies (⩾1 : 20) to C. pneumoniae; and 4-fold increases in antibody titer, with final titers for M. pneumoniae of ⩾1 : 160, for C. psittaci of ⩾1 : 256, for C. pneumoniae of ⩾1 : 512, for C. burnetii of ⩾1 : 160, and for L. pneumophila serogroups 1–6, respiratory syncytial virus, influenza A virus, and parainfluenza 3 virus indicative of seroconversion.

Etiologic diagnosis was considered presumptive when a predominant microorganism was isolated from a purulent sample (presence of >25 polymorphonuclear leukocytes and <10 squamous cells per low-magnification field [×10]) and the findings of Gram staining were compatible. Presumptive aspiration pneumonia was diagnosed on the basis of clinical and radiological findings for patients who had a predisposing cause of aspiration (compromised consciousness, altered gag reflex, or dysphagia) and radiographic evidence of involvement of a dependent pulmonary segment. Cases that did not meet any of the above etiologic diagnosis criteria were considered to be of unknown etiology.

To stratify patients in risk classes, we used the validated prediction rule calculated according to the PSI scores, developed by the Pneumonia Outcomes Research Team, as described elsewhere [12].

Statistical analysis. To detect significant differences between specified groups, we used the χ2 test with continuity correction for categorical variables and the Student's t test for continuous variables. For multiple comparisons, an analysis of variances with Bonferroni's correction was performed. Associations were considered statistically significant if the P value was <.05 in a 2-sided test. Multivariate analysis was performed by use of the stepwise logistic-regression model of the SPSS software package (SPSS).

Previous SectionNext Section

Results

Patients' characteristics. During the study period, a total of 533 nonimmunosuppressed adult patients with CAP were admitted to our institution, of whom 371 were men and 162 were women (mean patient age, 64 years; range, 16–96 years). Three hundred fifty-eight patients (67%) had underlying diseases, mainly chronic obstructive pulmonary disease (122 patients), diabetes mellitus (84), ischemic heart disease (46), chronic liver disease (34), congestive heart failure (29), solid tumor (26), chronic renal failure (20), HIV infection (18), and hematologic malignancies (8). One hundred fifty-four patients (29%) were smokers, 113 (21%) had a history of alcohol abuse, and 8 (2%) were injection drug users. Six patients (1%) were admitted to the hospital from a nursing facility. Two hundred sixteen patients (40%) had been immunized with influenza vaccine for the influenza season, and 19 patients (4%) had received 23-valent pneumococcal capsular polysaccharide vaccine in the 5 years prior to admission to the hospital. One hundred forty-eight patients (27%) had received antibiotic therapy before hospitalization. Radiography revealed evidence of lobar in 281 patients (53%), multilobar in 177 (33%), segmental in 67 (13%), and interstitial pneumonia in 8 (1%). Pleural effusion was observed in 107 patients (20%) and cavitation was observed in 15 patients (3%).

Etiology. Etiologic diagnoses were established in 283 cases (53.1%), of which 170 (60.1%) were classified as definitive and 113 (39.9%) as presumptive. More than 1 causative agent was found in 19 of the 283 cases of infection (6.7%). As shown in table 1, the most frequently identified pathogens were S. pneumoniae (in 135 cases), L. pneumophila (in 35), and Haemophilus influenzae (in 34).

Table 1
View larger version:
Table 1

Etiology of community-acquired pneumonia (CAP) in 533 patients who were hospitalized at the Bellvitge Hospital, Barcelona, Spain, from February 1995 through May 1997.

Overall, 51 patients had bacteremia (due to S. pneumoniae in 41 patients; H. influenzae, in 4; viridans group streptococci, in 4; Pseudomonas aeruginosa, in 1; and Porphyromonas asaccharolytica, in 1).

Risk classes and reasons for admission. PSI classes were as follows: class I, 51 patients; class II, 62; class III, 117; class IV, 198; and class V, 105. Table 2 shows the correlation of the conventional criteria used for hospitalization with the patients' PSI class. Two hundred thirty (43%) of our patients were in low-risk classes (I to III; PSI ⩽90); of these patients, 137 (60%) were hospitalized for the following reasons: respiratory failure (127 patients), large or complicated pleural effusion (17), and unstable vital signs (1). Of the remaining 93 patients (40%), 27 were hospitalized because of underlying disease, 24 for advanced age, 18 for lack of response to previous antibiotic therapy, 16 for multilobar pneumonia, 15 for clinical suspicion of a high-risk pathogen, and 2 for aspiration pneumonia. As shown in table 2, for some patients there was >1 criterion for admission. The main reasons for admitting low-risk patients to the hospital were the presence of comorbidities (in 54 [48%] of 113 patients) for patients in PSI risk classes I and II and respiratory failure (in 76 [65%] of 117) for patients in PSI risk class III.

Table 2
View larger version:
Table 2

Correlation of conventional criteria for hospitalization with the Pneumonia Severity Index (PSI) risk class.

Diagnostic workups were homogeneous among risk classes (table 3): no significant differences were found in the proportion of diagnostic techniques applied in patients in different risk classes. Overall, paired serological tests were performed more frequently for patients in low-risk classes (I to III) than they were for those in high-risk classes (P = .005), but when we compared only patients who survived long enough to have a second sample obtained, no significant differences were observed (P = .916). The distribution of principal pathogens according to risk classes is shown in table 4. The 41 patients with bacteremic pneumococcal pneumonia were classified as follows: class I, 6 patients; class II, 1; class III, 6; class IV, 10; and class V, 18.

Table 3
View larger version:
Table 3

Diagnostic workup and yield of different techniques, per Pneumonia Severity Index (PSI) risk class.

Table 4
View larger version:
Table 4

Etiologies and mortality rates according to the Pneumonia Severity Index (PSI) risk class in 533 patients who were hospitalized for community-acquired pneumonia at the Bellvitge Hospital, Barcelona, Spain, from February 1995 through May 1997.

Outcomes. All patients received empirical antibiotic therapy, which included β-lactams for 455 patients (85%), macrolides for 135 (25%), rifampin for 41 (8%), and another drug for 23 (4%). One hundred nine patients (20%) received more than 1 drug at admission to the hospital.

Mean time to resolution of symptoms was 6 days for patients with respiratory symptoms, 5 days for patients with chest pain, and 3 days for patients with fever. No significant differences were observed between patients in different risk classes with regard to time to resolution of symptoms.

Table 5 shows outcomes according to risk classes. Patients were hospitalized for a mean duration of 11 days. Duration of hospital stay was significantly longer for patients in class V than it was for patients in classes I to III. Complications during the hospital stay were observed in 148 patients (28%), the most frequent being empyema or pleural effusion (in 31 patients), heart failure (in 25), nosocomial infection (in 13), acute renal failure (in 12), gastrointestinal hemorrhage (in 8), diabetic coma (in 7), and hepatitis (in 7). Septic complications other than empyema were infrequent, and all of these complications were due to S. pneumoniae (meningitis in 2 patients; endocarditis, in 1; septic arthritis, in 1; and purulent pericarditis, in 1). The most common complications among patients in risk class I were empyema or pleural effusion (in 7 patients); in risk class II, arrhythmias (in 2); and in risk class III, empyema or pleural effusion (in 11) and heart failure (in 6).

Table 5
View larger version:
Table 5

Outcomes 533 patients with community-acquired pneumonia who were hospitalized at the Bellvitge Hospital, Barcelona, Spain, from February 1995 through May 1997, according to Pneumonia Severity Index (PSI) risk class.

Fifty-four patients (10%) died. Mortality rates according to PSI risk class were as follows: class I, 0%; class II, 2%; class III, 3%; class IV, 10%; and class V, 29%. Findings at admission to the hospital of a history of neoplasia, respiratory failure, renal failure, decreased level of consciousness, shock, and a simplified acute physiology score of >10 were independent risk factors for death (table 6). Mortality rates varied according to etiology, ranging from 0% in patients with viral pneumonia to 50% in patients with gram-negative pneumonia (table 4). The mortality rate associated with bacteremic pneumococcal pneumonia was 22%.

Table 6
View larger version:
Table 6

Factors at admission associated with mortality by univariate and multivariate analysis for 533 patients with community-acquired pneumonia who were hospitalized at Bellvitge Hospital, Barcelona, Spain, from February 1995 through May 1997.

Previous SectionNext Section

Discussion

The present study offers an exhaustive clinical evaluation of a series of patients hospitalized for CAP in a single institution and places special emphasis on the etiologic diagnosis, reasons for admission, and outcomes. All decisions about the occurrence of medical complications and final diagnostic classifications were made by consensus among the clinical investigators during the study.

Baseline clinical characteristics of patients with regard to demographics, underlying diseases, and prior antibiotic use were similar to those reported by others [16–18]. Our series did not include immunosuppressed persons, and although a significant number of our patients were of advanced age, there were few patients admitted to the hospital from a nursing facility.

As in most other series of hospitalized patients with CAP, S. pneumoniae was the leading causative organism [16–21]. Legionella, H. influenzae, and aspiration pneumonia were also among the most frequently identified etiologies. The high number of L. pneumophila diagnoses in our series is remarkable. It should be noted that most cases were diagnosed on the basis of detection of urinary antigen and/or culture of respiratory specimens, and that we considered the results of serological tests to be positive only when a 4-fold increase in IgG titer was documented.

On the other hand, C. pneumoniae was infrequently a cause of pneumonia in our hospitalized population; this may be in part because we also disregarded cases in which single or persistently high IgG titers were the only diagnostic finding. In fact, 45 patients (8%) had this serological profile (IgG, ⩾1 : 512), and for 25 (56%) of them, other etiologic diagnoses were established. The inclusion of these cases as true C. pneumoniae-related episodes would have increased the number of cases of chlamydia pneumonia from 11 to 56 and the percentage of cases of mixed pneumonia to 8%. Because the clinical significance of a single high titer of IgG to C. pneumoniae is still unclear, we decided to consider this diagnosis only for cases in patients in whom seroconversion was evident [19, 22–25].

A finding in our study worthy of particular attention is the correlation between different etiologies and the PSI classes, especially since diagnostic workups were homogenous among risk classes. As expected, S. pneumoniae was the most frequently identified causative organism among patients in all risk classes, although it was found more frequently in patients in high-risk classes, as were L. pneumophila and H. influenzae. On the other hand, patients with atypical pneumonia as a sole diagnosis (excluding Legionella species) were predominantly in low-risk classes, whereas those with gram-negative, aspiration, and bacteremic pneumococcal pneumonia and those with mixed infections (typical plus atypical) were usually assigned to high-risk classes. It is interesting that there were no deaths among patients with mixed infections, a finding that suggests that their assignment to class IV or V may have been due mainly to advanced age and the presence of comorbidities.

Mortality rates were particularly high in association with bacteremic pneumococcal, gram-negative, and aspiration pneumonia, an observation that has been reported elsewhere [5, 26]. Patients with these high-risk etiologies were indirectly identified by the PSI score, because they were more often assigned to high-risk classes. Factors at admission that were associated with higher mortality rates were similar to those identified in other series; as expected, the PSI accurately identified patients at risk of dying, and our figures were very close to those reported by Fine et al. [12].

A striking finding was that a high proportion of patients were assigned to low-risk classes—not only class III (117 patients [22%]), but also classes I (51 [10%]) and II (62 [12%]). This suggests that for patients with PSI scores of ⩽90, the decisions to hospitalize that were based on conventional criteria were not in concordance with the prediction rule. Indeed, Fine et al. [12] stated that the rule should never supersede clinical judgement. Nevertheless, the widespread implementation of the rule in different institutions and countries may lead to the idea that no low-risk patient should be hospitalized; of particular concern is its “strict” use to decide the site of care. In a pilot study, Atlas et al. [6] showed that the implementation of PSI scores to determine hospital admission practices for a cohort of patients assigned to groups I to III significantly reduced admission rates in comparison with the admission rates for a historical group. However, readmission of patients who experienced failure of outpatient management occurred more frequently than had been expected, and patient satisfaction did not increase with use of the new approach.

Several factors that the rule does not take into account may greatly influence the decision to hospitalize [27]. Apart from factors related to the patient, such as poor social support and inability to maintain oral intake, the most frequently recognized reason for admission to the hospital is the need for supplementary oxygen. In our study, we found that patients in classes I to III were hospitalized for a wide range of reasons, often multiple, including comorbidities, respiratory failure, and complications such as clinically significant pleural effusion.

In a further analysis of these discrepancies, 2 different groups of patients appear. The first group—who presented with respiratory failure and a need for supplementary oxygen, with unstable vital signs, or with large or complicated pleural effusion—should indisputably be hospitalized and do account for 60% of the low-risk class patients. The relevance of pleural effusion, which has not been previously recognized as a requirement for hospitalization in implementation of the rule, should be stressed. The second group of patients (i.e., the remaining 40%) were hospitalized because of other factors such as advanced age and presence of uncomplicated comorbidities. It is possible that some of these patients could have been treated as outpatients from the beginning. The most appropriate treatment of these patients deserves further consideration in controlled studies. In this regard, the group of patients in classes II and III, who are without significant pleural effusion and not in need of supplementary oxygen, seem particularly suited to outpatient treatment, especially because outpatient treatment of low-risk patients is associated with more rapid resumption of normal activities [28].

It is clear that outcome measures other than death, such as need of subsequent hospitalizations, pleural complications, adverse drug reactions, restoration of normal activities, and use of health resources, should be considered when the site of care is decided, but evidence-based data are still lacking. In addition, other variables, such as cost, quality of life, and patient preference, should also be considered. Because a significant number of patients who are currently hospitalized on the basis of classic criteria are in low-risk PSI classes, and because the reasons for their admission to the hospital are not irrefutable, prospective randomized studies are needed to determine outcomes other than death among these patients and to compare ambulatory and in-hospital management.

Previous SectionNext Section

Acknowledgments

We thank the staff and residents of the infectious disease, respiratory, and microbiology services for their valuable cooperation.

Previous SectionNext Section

Footnotes

  • This prospective longitudinal and observational study was approved by the Ethical Committee at our institution.

  • Financial support: Fondo de Investigaciones Sanitarias de la Seguridad Social (95/1100) and (via grants to Dr. Rosón) Ciutat Sanitària i Universitària de Bellvitge (1995), Fondo de Investigaciones Sanitarias de la Seguridad Social (96/5163 and 97/5245), and Fundació Universitària Agustí Pedro i Pons (1998).

  • Received June 27, 2000.
  • Revision received November 14, 2000.
Previous Section
 

References

  1. ↵
    1. Marrie TJ
    . Community-acquired pneumonia. Clin Infect Dis 1994;18:501-15.
    FREE Full Text
    1. Barlett JG,
    2. Mundy LM
    . Community-acquired pneumonia. N Engl J Med 1995;333:1618-24.
    CrossRefMedlineWeb of Science
  3. ↵
    1. Marrie TJ
    . Management decisions regarding community-acquired pneumonia. Curr Opin Infect Dis 1999;12:133-6.
    MedlineWeb of Science
  4. ↵
    1. Dean NC
    . Use of prognostic scoring and outcome assessment tools in the admission decision for community-acquired pneumonia. Clin Chest Med 1999;20:521-9.
    CrossRefMedlineWeb of Science
  5. ↵
    1. Fine MJ,
    2. Smith DN,
    3. Singer DE
    . Hospitalization decision in patients with community-acquired pneumonia: a prospective cohort study. Am J Med 1990;89:713-21.
    CrossRefMedlineWeb of Science
  6. ↵
    1. Atlas SJ,
    2. Benzer TI,
    3. Borowsky LH,
    4. et al
    . Safely increasing the proportion of patients with community-acquired pneumonia treated as outpatients. Arch Intern Med 1998;158:1350-6.
    Abstract/FREE Full Text
  7. ↵
    1. Fine MJ,
    2. Hough LJ,
    3. Medsger AR,
    4. et al
    . The hospital admission decision for patients with community-acquired pneumonia. Arch Intern Med 1997;157:36-44.
    Abstract/FREE Full Text
  8. ↵
    British Thoracic Society. Community-acquired pneumonia in adults in British hospitals in 1982–1983: a survey of aetiology, mortality, prognostic factors and outcome. Q J Med 1987;62:195-222.
    Abstract/FREE Full Text
    1. Farr BM,
    2. Sloman AJ,
    3. Fisch MJ
    . Predicting death in patients hospitalized for community-acquired pneumonia. Ann Intern Med 1991;115:428-36.
    Abstract/FREE Full Text
  10. The ATS Board of Directors. Guidelines for the initial management of adults with community-acquired pneumonia: diagnosis, assessment of severity, and initial antimicrobial therapy. Am Rev Respir Dis 1993;148:1418-26.
    MedlineWeb of Science
  11. ↵
    1. Fine MJ,
    2. Smith MA,
    3. Carson CA,
    4. et al
    . Prognosis and outcomes of patients with community-acquired pneumonia. JAMA 1996;275:134-41.
    Abstract/FREE Full Text
  12. ↵
    1. Fine MJ,
    2. Auble TE,
    3. Yealy DM,
    4. et al
    . A prediction rule to identify low-risk patients with community-acquired pneumonia. N Engl J Med 1997;336:243-50.
    CrossRefMedlineWeb of Science
  13. ↵
    1. Farr BM
    . Prognosis and decisions in pneumonia. N Engl J Med 1997;336:288-9.
    CrossRefMedlineWeb of Science
  14. ↵
    1. García A,
    2. Rosón B,
    3. Pérez JL,
    4. et al
    . Usefulness of PCR and antigen latex agglutination test with samples obtained by transthoracic needle aspiration for diagnosis of pneumococcal pneumonia. J Clin Microbiol 1999;37:709-14.
    Abstract/FREE Full Text
  15. ↵
    1. Le Gall JR,
    2. Loirat P,
    3. Alpérovitch A,
    4. et al
    . A simplified acute physiology score for ICU patients. Crit Care Med 1984;12:975-7.
    MedlineWeb of Science
  16. ↵
    1. Marrie TJ,
    2. Durant H,
    3. Yates L
    . Community-acquired pneumonia requiring hospitalization: 5-year prospective study. Rev Infect Dis 1989;11:586-99.
    MedlineWeb of Science
    1. Mundy LM,
    2. Auwaerter PG,
    3. Oldach D,
    4. et al
    . Community-acquired pneumonia: impact of immune status. Am J Respir Crit Care Med 1995;152:1309-15.
    Abstract/FREE Full Text
  18. ↵
    1. Fang GD,
    2. Fine M,
    3. Orloff J,
    4. et al
    . New and emerging etiologies for community-acquired pneumonia with implications for therapy: a prospective multicenter study of 359 cases. Medicine (Baltimore) 1990;69:307-16.
    Medline
  19. ↵
    1. Mundy LM,
    2. Oldach D,
    3. Auwaerter PG,
    4. et al
    . Implications for macrolide treatment in community-acquired pneumonia. Chest 1998;113:1201-6.
    Abstract/FREE Full Text
    1. Marston BJ,
    2. Plouffe JF,
    3. File TM,
    4. et al
    . The Community-Based Pneumonia Incidence Study. Arch Intern Med 1997;157:1709-18.
    Abstract/FREE Full Text
  21. ↵
    1. Sopena N,
    2. Sabrià M,
    3. Pedro-Botet ML,
    4. et al
    . Prospective study of community-acquired pneumonia of bacterial etiology in adults. Eur J Clin Microbiol Infect Dis 1999;18:852-8.
    CrossRefMedlineWeb of Science
  22. ↵
    1. Gaydos CA,
    2. Roblin PM,
    3. Hammerschlag MR,
    4. et al
    . Diagnostic utility of PCR-enzyme immunoassay, culture, and serology for detection of Chlamydia pneumoniae in symptomatic and asymptomatic patients. J Clin Microbiol 1994;32:903-5.
    Abstract/FREE Full Text
    1. Lévy M,
    2. Dromer F,
    3. Brion N,
    4. Leturdu F,
    5. Carbon C
    . Community-acquired pneumonia: importance of initial noninvasive bacteriologic and radiographic investigations. Chest 1988;93:43-8.
    Abstract/FREE Full Text
    1. Kern DG,
    2. Neill MA,
    3. Schacter J
    . A seroepidemiologic study of Chlamydia pneumoniae infection in Rhode Island: evidence of serologic cross-reactivity. Chest 1993;104:208-13.
    Abstract/FREE Full Text
  25. ↵
    1. Petitjean J,
    2. Vincent F,
    3. Fretigny M,
    4. et al
    . Comparison of two serological methods and a polymerase chain reaction-enzyme immunoassay for the diagnosis of acute respiratory infections with Chlamydia pneumoniae in adults. J Med Microbiol 1998;47:615-21.
    Abstract/FREE Full Text
  26. ↵
    1. Ruiz M,
    2. Ewig S,
    3. Torres A,
    4. et al
    . Severe community-acquired pneumonia: risk factors and follow up epidemiology. Am J Respir Crit Care Med 1999;160:923-9.
    Abstract/FREE Full Text
  27. ↵
    1. Halm EA,
    2. Atlas SJ,
    3. Borowsky LH,
    4. et al
    . Understanding physician adherence with a pneumonia practice guideline: effects of patient, system and physician factors. Arch Intern Med 2000;160:98-104.
    Abstract/FREE Full Text
  28. ↵
    1. Pomilla PV,
    2. Brown RB
    . Outpatient treatment of community-acquired pneumonia in adults. Arch Intern Med 1994;154:1793-802.
    Abstract/FREE Full Text
| Table of Contents

This Article

  1. Clin Infect Dis. (2001) 33 (2): 158-165. doi: 10.1086/321808
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 54 (5)
  1. Clinical Infectious Diseases
  1. Alert me to new issues

Most