0Original Report
Prospective epidemiologic survey of patients with community-acquired pneumonia requiring hospitalization in Switzerland
J. Garbino/1) R. Sommer,(l) A. Gerber/2) C. Regamey/3) P. Vernazza,(4) D. Genne,'5» P. Diir,(6) M. Rothen/7) J.-P. Unger'8' and D. Lew0'
Background: Community-acquired pneumonia (CAP) is a common problem and the principal infection requiring hospitalization, but its treatment is complicated by the difficulty in microbiological diagnosis and the increasing incidence of antibiotic resistance among respiratory pathogens. The purpose of this paper is to present the main epidemiologic features of patients with CAP requiring hospitalization in our country.
Methods: We enrolled three hundred and eighteen adult patients with CAP requiring hospitalization in seven large medical centers in Switzerland during two winter periods. The patients' mean age was 70.4 years. This study describes the epidemiology of these patients. Clinical, radiologic and microbiological evaluations were performed at study entry during treatment, and at 4 weeks post-therapy. For microbiological diagnostic purposes, sputum culture, throat swab culture, PCR, blood cultures, Legionella urinary antigen and serologic evaluations were also performed.
Results: Despite the higher mean age, the overall mortality rate was 8%, lower than in other comparable studies. The most common underlying diseases present at study entry were cardiac failure (23%), chronic obstructive pulmonary disease (20%), renal failure (15%), and diabetes (12%);40% of the patients were smokers. Although dyspnea, cough and positive pulmonary auscultation findings were present in about 90% of patients, fever >38°C was present in only 64%. The most frequently isolated respiratory pathogens were Streptococcus pneumoniae (12.6%), Haemophilus influenzae (6%), Staphylococcus aureus (1.6%), and Moraxella catarrhalis (1.6%). Atypical pathogens were frequently found, with the following distribution: Mycoplasma pneumoniae, 7.5%; Chlamydia pneumoniae, 5.3%; and Legionella pneumophila, 4.4%. The mean duration between onset of symptoms and hospital admission was 4.8 days, and the mean treatment duration was 12.1 days. Two weeks after the start of therapy, although clinical symptoms were absent, radiologic infiltrates were still present in 24% of patients.
Conclusions: The microbiological diagnosis in CAP can be established in only about 50% of cases with the combination of several diagnostic tools. Epidemiologic surveys of CAP should be performed on a regular basis, regionally, as a way to improve the management of these infections.
Int J Infect Dis 2002; 6: 288-293
INTRODUCTION
Respiratory infections are among the illnesses most frequently encountered by physicians in the USA and
(''Division of Infectious Diseases, Department of Internal Medicine, University Hospitals of Geneva, Geneva, Switzerland; ,2lDepartment of Internal Medicine, Regional Hospital, Biel, Switzerland; <3)Depart-ment of Internal Medicine, Hópital Cantonal, Fribourg, Switzerland; <4)Department of Internal Medicine, Kantonsspital, Saint Gall, Switzerland; (5)Department of Internal Medicine, Hópital de la ville, La Chaux des Fonds, Switzerland; ''"Department of Internal Medicine, Kantonsspital, Zoug, Switzerland; '''Department of Internal Medicine, Kantonsspital, Bruderholz, Switzerland; (8)Division of Emergency Medicine, Department of Internal Medicine, University Hospitals of Geneva, Geneva, Switzerland. Supported by Pfizer Inc., Switzerland.
Address correspondence to Jorge Garbino, Division of Infectious Diseases, Department of Internal Medicine, University Hospitals of Geneva, 24, rue Micheli-du-Crest, 1211 Geneva 14, Switzerland. E-mail: jorge.garbino@hcuge.ch
Corresponding Editor: Patricia Muñoz, Madrid, Spain
Europe in both outpatient and hospital practice. In particular, community-acquired pneumonia (CAP) has an estimated annual incidence of 4 million cases among ambulatory patients, or 12 cases per 1000 persons in the USA. Pneumonia, in general, is the sixth leading cause of death in the USA, and is responsible for more than 600 000 hospital admissions per year.1 Mortality rates approximate 13.7% among all cases of CAP requiring hospital admission,2 and appear to be even higher in the elderly population.
The most common microbes associated with CAP are Streptococcus pneumoniae, Haemophilus influenzae, and Mycoplasma pneumoniae? 5 However, during the last decade, pathogens such as Moraxella catarrhalis, Legionella pneumophila and Chlamydia pneumoniae have been identified more frequently, and several studies have demonstrated the importance of these newly recognized pathogens.2
Diagnosis of infections caused by atypical pathogens is often limited by the lack of routine and easy-to-perform tests for these agents. In addition, the growing
incidence of (3-lactam and/or macrolide resistance in common respiratory bacterial pathogens such as Streptococcus pneumoniae and Haemophilus influenzae complicates the selection of an appropriate antibiotic for initial empirical treatment.6"9 Initial empirical therapy should inhibit Streptococcus pneumoniae, Haemophilus influenzae, Moraxella catarrhalis, and atypical pathogens.
The initial management of CAP requires a decision as to the need for patient hospitalization. In selected circumstances, the severity of illness10 will necessitate hospital admission and the administration of intravenous antibiotics. Once the patient has improved, intravenous antibiotics may be replaced with oral agents, depending on the availability of an efficacious and safe oral agent to treat the infecting pathogen(s).
We have performed a prospective, randomized, open-label, comparative, controlled study on CAP requiring hospitalization in Switzerland.
The original study was a randomized, phase III, multicenter, blind (open when patient was included) study to compare the quinolone trovafloxacin 200 mg/day intravenously, and then orally once a day, with amoxi-cillin-clavulanic acid 4 x 1.2 g/day intravenously, and then orally with optional clarithromycin, for the treatment of CAP. The intravenous treatment was given for a minimum of 48 h, followed by oral treatment, to complete at a course of treatment of at least 7 days.
The purpose of this paper is to present the main epidemiologic features of patients with CAP requiring hospitalization in our country, and not to compare the results of the two different treatments.
MATERIALS AND METHODS
The study was a multicenter, prospective, randomized, blind (open-label when patient was included), comparative, controlled study conducted at seven Swiss medical centers (Basel-Bruderholz, Biel, Fribourg, Geneva, La Chaux-de-Fonds, Saint-Gall, and Zoug). The institutional review board of each center approved the study, and all patients provided written informed consent.
Patients who fulfilled inclusion criteria were randomized according to a computer-generated, blinded randomization list, and assigned to one treatment arm to complete at least 7 days of therapy.
Patient characteristics
Patients eligible for inclusion in the study met the following criteria: older than 18 years and hospitalized with a primary diagnosis of CAP; a medical and clinical history of pneumonia; a new pulmonary infiltrate revealed by chest X-ray compatible with pneumonia; and two or more signs and symptoms associated with a lower respiratory tract infection (i.e. body temperature >38°C or <36.1°C, new or increased cough, production of purulent sputum, findings on examination such as rales or evidence of pulmonary consolidation, blood
leukocyte count >10 000/mm3 or <4000/mm3, or >10% bands, hypoxemia (defined as:p02 <70 mmHg at Fi02= 21% or a decrease of >25% from an initial value)).
Patients excluded were: pregnant women, nursing mothers, or women of childbearing potential not practicing adequate contraception; patients with known or suspected hypersensitivity or intolerance to the study drugs; and those who had received antimicrobial therapy for 24 h or longer within 72 h prior to the baseline visit, unless there was documented evidence of resistance or clinical failure. Also excluded were the following subjects: those with AIDS, suspected Pneumocytis carinii pneumonia, or a known infection with fungi or mycobacteria; those with neutropenia, defined as a total of white blood cells of less than 1000 leukocytes/mm3, or an absolute neutrophil count of less than 1000/mm3; those receiving immunosuppressive therapy, defined as chronic treatment with known immunosuppressive agents; those with a diagnosis of cystic fibrosis, known lung cancer, emphysema or aspiration pneumonia, cavitary lung diseases confirmed by chest X-ray at baseline, or a history of seizures; those who had received any investigational agent within the month prior to the baseline visit; those who had been enrolled in this study previously; those with evidence of recent drug or alcohol abuse or dependence; and those with an underlying condition that might give them a life expectancy <1 week, or who had been hospitalized during the past week.
Microbiological investigations
From'all patients, a sputum specimen for routine culture and Gram stain was obtained prior to study entry. If a satisfactory specimen or expectorated sputum was not obtained, the investigator could carry out induction of sputum with nebulized hypertonic or physiotherapy or, if medically indicated, could use other techniques such as transtracheal aspiration, bronchoalveolar lavage, bronchial brushings, or biopsy. In addition, blood cultures were performed by standard procedures at study entry and during other visits, if clinically indicated. Identification of causative organisms and testing for aerobic susceptibility to study drugs by disk and broth microdilution MIC methods was performed according to the National Committee for Clinical Laboratory Standards procedures.11 Serologic testing for Mycoplasma pneumoniae (complement fixation), L. pneumophila (enzyme-linked immunosorbent assay) and Chlamydia pneumoniae (microimmunofiuorescence test) was performed in all patients at baseline and 4 weeks after study entry (results were positive when there was a fourfold rise in the titer of Ig or the Ig titer was >1/256 in a simple serum sample). A urine antigen assay for Legionella spp. was also performed at the time of admission (Binax, S. Portland, Maine, USA).12
Identification of typical and atypical pathogens was performed at the bacteriologic laboratories of the indicated centers, and, when necessary, by a reference
290 International Journal of Infectious Diseases / Volume 6, Number 4,2002
laboratory. Standard respiratory pathogens, when isolated from purulent respiratory secretions, were considered to be the microorganism responsible for the infection. Throat swabs and sputum specimens were used at study entry for PCR assays for Mycoplasma pneumoniae (gene amplification with primers MPP10 and MPP12)13 and Chlamydia pneumoniae (amplification by Pst with primers HM-1 and HR-1).14
RESULTS
Three hundred and twenty-three patients were enrolled during the 19-month study period (from November 1997 to May 1999) at seven Swiss hospital centers. Five patients did not meet the entry criteria and were excluded from the study due to protocol violations (hospital-acquired pneumonia, bronchial aspiration pneumonia, acute bronchitis, and two cases of urinary sepsis). The numbers of patients included in the different centers were as follows: Basel-Bruderholz, 6; Biel, 60; Fribourg, 51; Geneva, 102; La Chaux-de-Fonds, 42; Saint Gall, 47; and Zoug, 10. The total number of patients was 318.
Study population
The mean age of the adult patients was 70.4 years. Baseline demographic data are presented in Table 1.
The most frequent underlying diseases present at study entry were cardiac failure (71, 23%), chronic obstructive pulmonary disease (63, 20%), renal failure (45, 15%), diabetes (39, 12%), asthma (17, 5%), and hepatic failure(16, 5%). One hundred and twenty-four patients (40%) were smokers. The mean time from onset of symptoms to admission was 4.8 days (SD 4.39) (median 3.5). The mean total duration of treatment for clinically evaluable patients was 12.1 days (±4.95).
The proportions of patients with different clinical symptoms and signs of pneumonia at study entry are presented in Table 1.
The relatively low frequency of fever was probably due to the advanced age of the patients. The high mean age of the study population and the high percentage of co-morbidities allow the classification of patients as class III according to the Fine criteria.15
Radiologic results
The most common localization of radiologic infiltrates at study entry comprised lesions on the right lung (49%), with a significant difference (P<0.05) compared to localization on the left lung (34%). Bilateral infiltrates were found in 81 patients (18%). Pleural effusion was present in 86 patients (27%). Other radiologic data are shown in Table 2.
Two weeks after the start of treatment, radiologic success (defined as resolution or improvement of abnormal pre-therapy radiographic findings) was observed in 188 patients (76%), whereas radiologic failure (failure
of resolution or improvement of abnormal pre-therapy radiographic findings) was observed at the same time in 62 patients (24%).
At the end of the study period (28 days after study entry), among evaluable patients, radiologic resolution (no abnormal radiographic findings) was observed in 102, and persistence of radiologic infiltrates in 13.
Bacteriologic findings
One hundred and twenty-four pathogens (39%) were identified at baseline among the patients. The most common typical bacterial pathogens isolated from sputum specimens of the study population patients were Streptococcus pneumoniae (40), Haemophilus influenzae (19), Staphylococcus aureus (8), and Moraxella catarrhalis (5). Other microorganisms isolated from sputum cultures are listed in Table 3. For 28 patients, sputum could not
Table 1. Patients' demographic characteristics and clinical symptoms and signs of pneumonia at study entry
Demographic
Total number of patients 318
Male (no. (%)) 188 (59%)
Female (no. (%)) 130 (41%)
Age (years) 70.4 (±16.2)
Patients over 65 years (no. (%)) 212 (67%)
Body weight (kg) 80 (37-123)
Height (cm) 168.5 (145-195)
Signs and symptoms (no. (%))
Pulmonary auscultation findings 291 (92%)
Cough 287 (90%)
Dyspnea 280 (88%)
Respiratory rate
>20/min 249 (78%)
>30/min 95 (30%)
Increased sputum volume 215 (68%)
Temperature >38°C 202 (64%)
Heart rate >95/min 173 (54%)
Chills 173 (54%)
Pleuritic chest pain 168 (53%)
Pleural effusion 86 (27%)
Table 2. Radiologic findings in the study population at study
Number %
Presence of lung infiltrates (n=478)
In right lung 235 49
In left lung 162 34
Bilateral lung 81 17
In right lung lobes (n=308)
Lower 138 45
Middle 77 25
•Upper 55 18
Bi-lobe 34 11
Tri-I6be 4 1
In left lung lobes (n = 170)
Lower 126 74
Upper 31 18
Both 13 8
Table 3. Respiratory tract pathogens detected by sputum culture and nonculture methods in the study population
Number of
Microorganisms patients Percentage
Streptococcus pneumoniae 40 12.6
Haemophilus influenzae 19 6.0
Mycoplasma pneumoniae 24 7.5
Chlamydia pneumoniae 8 2.5
Legionella spp. 17 5.3
Staphylococcus aureus 14 4.4
Moraxella catarrhalis 5 1.6
Klebsiella pneumoniae 2 0.6
Streptococcus beta-haemolyticus 1 0.3
Mycobacterium (non-tuberculosis) 1 0.3
Pasteurella multocida 1 0.3
For L. pneumophila, seven serology conversion, six cases diagnosed by positive urinary antigen, and one by culture. For Mycoplasma pneumoniae, seven serology conversion and 17 diagnosed by PCR.
For Chlamydia pneumoniae, nine serology conversion and eight diagnosed by PCR.
be obtained at study entry. No penicillin-, erythromycin-or clarithromycin-resistant isolates of Streptococcus pneumoniae were identified.
Other microorganisms isolated from sputum cultures at study entry which are not usually considered to be respiratory pathogens were: Pseudomonas aeruginosa, Escherichia coli, Enterobacter cloacae, Acinetobacter spp., Corynebacterium diphtheriae, Pseudomonas fluorescens, Stenotrophomonas maltophilia, Serratia marcescens, Enterobacter aerogenes, Hafnia alvei, Candida albicans, and Aspergillus fumigatus. None of the patients was subjected to invasive techniques to obtain sputum samples.
Serologic samples were taken from 83% (264) of the patients at study entry, and from 53% (169) patients 4 weeks later for serology control.
Legionella urinary antigen testing was performed in 95% of the patients at study entry, and was positive in six cases. Positive PCR in throat specimens was observed in 17 cases for Mycoplasma pneumoniae, and in eight cases for Chlamydia pneumoniae. An etiologic diagnosis of CAP was possible in 130 patients (41%).The overall mortality rate at 30 days in this patient population was 8%.
Blood cultures
At study entry, 56 (19.2%) patients had positive blood cultures. The most common bacterial pathogens isolated from blood cultures and presumed to be involved in the etiology of pneumonia were: Streptococcus pneumoniae (n=32) (10%) (57% of all positive blood cultures), Haemophilus influenzae, Staphylococcus aureus, Escherichia coli (n = 3) (5.3%), Morganella morganii, and Streptococcus agalactiae (n=l) (1.8%). Other microorganisms isolated without evidence of a pathogenic role in pneumonia or considered to be contaminants were coagulase-negative Staphylococcus (n=7), Propionibacterium spp. (n=2) and, with one case each, Enterococcus aerogenes, Pseudomonas pancinoli, Streptococcus mitis, and Oligella urethralis.
Analysis of all microorganisms isolated from blood cultures (56) and all microorganisms isolated from sputum (124) showed that matching of similar pathogens (blood cultures/sputum) was achieved in only nine cases. In 29 cases with positive blood cultures, no microorganisms were found in sputum. Finally, in 18 cases with positive blood cultures, the microorganisms isolated in sputum were different but did not fit the standard purulent criteria with Gram stain. These results demonstrate the importance of blood cultures for the establishment of the etiology of CAP, and give an indication of how frequently blood or sputum samples may contaminate pathogens.
DISCUSSION
CAP is a common condition, and its treatment poses new challenges because of the growing incidence of antibiotic resistance among traditional respiratory pathogens and the increasing recognition of pneumonia caused by atypical respiratory pathogens.16
Rapid diagnosis of the etiologic agents causing respiratory infections is difficult, and the treatment of patients with CAP is often empirical. Therefore, selection of an appropriate antibacterial agent should be, in large part, based on its antimicrobial activity and the local epidemiologic setting. Local sensitivity patterns need to be considered in particular, because there are different rates of pneumococci penicillin resistance in different countries.
In this study, we achieved a precise etiology for the causative pathogens, with the use of different diagnostic techniques such as sputum culture, PCR, serology, blood cultures, and Legionella urine antigen, in a total of 41 % of the study population. The etiology was unknown for the other patients, but this proportion is comparable with the results of other studies.217 Antibiotics given before admission probably decreased the possibility of isolating a specific pathogen, even though antibiotic treatment was administered for less than 24 h in our study.
The most frequent typical respiratory pathogens isolated were: Streptococcus pneumoniae (the predominant pathogen in our study, as in other published studies21819), Haemophilus influenzae, Staphylococcus aureus, Moraxella catarrhalis, and Klebsiella pneumoniae. Infection rates with the three atypical pathogens Chlamydia pneumoniae, Legionella pneumophila and Mycoplasma pneumoniae were comparable to those in the series published by Halm et al,20 but lower than in other series.121
In the present study, we did not observe antibiotic resistance of Streptococcus pneumoniae, as reported elsewhere.9 The resistance of Streptococcus pneumoniae in the Alexander project during the same period ranged between 3% and 6% in Switzerland (from 1996 to 1999). The absence of resistant pneumococci in our patients is
292 International Journal of Infectious Diseases I Volume 6,
probably due to differences between overall specimen collection in some large microbiological centers when compared to our prospective clinical study.
In some other European countries, such as Italy, Austria and Belgium, the frequency of resistance in Streptococcus pneumoniae remains quite low, but not in France or Greece, where it was much higher during the same time period (32-40%). In the study by Plouffe et al,22 from the USA, incidences of 8% of Streptococcus pneumoniae resistance to penicillin and 2.8% to ofloxacin were reported. We did not observe Streptococcus pneumoniae resistance to erythromycin and clarithromycin, whereas in another study, performed in Providence (USA), 5% of isolates of Streptococcus pneumoniae from bacteremic patients were resistant to erythromycin and clarithromycin.23
The most common localization of radiologic infiltrates comprised lesions on the right lung, with significant differences to localization on the left lung. Bilateral infiltrates were found in 81 patients. Radiologic resolution was not observed in 13 patients.
The studies listed in Table 4 were comparable to the present study. They were multicenter and randomized studies with similar sample sizes. Most of the studies were performed in North America; therefore, it is
Number 4,2002
interesting to have recent epidemiologic data from Switzerland. Some of these studies looked for concomitant diseases, clinical signs and clinical symptoms at study entry. In our patient population, the most important differences, compared to these other studies, were a higher patient mean age, a higher proportion of smokers, and a more accurate description of respiratory symptoms and clinical signs. In our series, careful auscultation revealed positive signs in 92% of the cases. By contrast, fever was observed in only 64% of cases.
The mortality rate of 8% among our hospitalized patients was lower than those previously reported in most other studies,2 despite the fact that the severity of disease and the age at hospitalization were high. This probably reflects a better standard of care, and perhaps earlier treatment in the ambulatory setting preventing a progressive severe infection. The microbiological diagnosis in CAP can only be established in about 50% of cases. In spite of the combined use of several diagnostic tools, about 50% of patients are treated with no identification of the microorganisms involved. We recommend that epidemiologic surveys of CAP should be performed on a regular basis, regionally, as a way to improve the management of these frequent infections requiring hospitalization.
Table 4. Comparison with other studies
Sarbino, Mandell and Halm Fang Plouffe File Bartlett
present report Donowitz24 et al20 et al2 et al2' et al'7 and Mundy1
Study design
Multicenter X X X X X
Randomized X X X X X
Mean age (years) 70 50-60 58 63 57-58 50
Number of patients 318 359 298 456
Etiology
Known etiology (%) 41 50-70 27 37 64 46 35-70
Positive blood cultures (%) 17.6 7.70 11
Streptococcus pneumoniae (%) 12 16-60 8 15.3 8 13.5 5-18
Haemophilus influenzae (%) 6 4-15 4 10.9 10 11.2 3-18
Gram-negative pathogens (%) 7.9 7-18 3 5.9 3-10
Atypical pathogens (%) 4 1-10 2 26 10-20
Concomitant diseases
COPD(%) 19.8 21 31.4
Cardiac failure (%) 22 19 29
Renal failure (%) 14.1 8 7
Diabetes (%) 12.2 13
Hepatic failure (%) 5
Smoker (%) 38.9
Asthma (%) 5 9
Symptoms (%) 45.1
Heart rate >95/min 55 45 56 80
: Temperature >38°C 64 68-78 46
Respiratory rate >25/min 78 69 71 47.2 49.4 >85
Cough 90 80 88.4
Dyspnea 88 60.2 57.8
Chills 54 40-50 47.8 49.4
Pleural effusion 27 20 30-50
Increased sputum production 68 .60-80 70.8 67.9 80
Pulmonary auscultation. 92
COPD, chronic obstructive pulmonary disease.
REFERENCES
1. Bartlett JG, Mundy LM. Community-acquired pneumonia [see comments], N Engl J Med 1995; 333(24):1618-1624.
2. Fang GD, Fine M, Orloff J, 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(5):307-316.
3. Fass RJ. Aetiology and treatment of community-acquired pneumonia in adults: an historical perspective. J Anti-microb Chemother 1993; 32(suppl A):17-27.
4. Berntsson E, Lagergard T, Strannegard O, Trollfors B. Etiology of community-acquired pneumonia in out-patients. Eur J Clin Microbiol 1986; 5(4):446-447.
5. Ball P. Therapy for pneumococcal infection at the millennium: doubts and certainties. Am J Med 1999; 107(1A): 77S-85S.
6. Breiman RF, Butler JC, Tenover FC, Elliott JA, Facklam RR. Emergence of drug-resistant pneumococcal infections in the United States [see comments], JAMA 1994; 271(23): 1831-1835.
7. Ashby BL. Treatment of pneumonia: new strategies for changing pathogens. ClinTher 1991; 13(5):637-650.
8. Friedland IR, McCracken GH Jr. Management of infections caused by antibiotic-resistant Streptococcus pneumoniae [see comments], N Engl J Med 1994; 331(6):377-382.
9. Jorgensen JH, Doern GV, Mäher LA, Howell AW, Redding JS. Antimicrobial resistance among respiratory isolates of Haemophilus influenzae, Moraxella catarrhalis, and Streptococcus pneumoniae in the United States. Antimicrob Agents Chemother 1990; 34(ll):2075-2080.
10. Ewig S, Torres A. Severe community-acquired pneumonia. Clin Chest Med 1999; 20(3):575-587.
11. National Committee for Clinical Laboratory Standards. Methods for dilution antimicrobial testing for bacteria, 5th edn. Approved standard. M7-A5/M100. Wayne, PA: NCCLS, 2000.
12. Plouffe JF, File TM Jr, Breiman RF, et al. Réévaluation of the definition of Legionnaires' disease: use of the urinary antigen assay. Community Based Pneumonia Incidence Study Group. Clin Infect Dis 1995; 20(5):1286-1291.
13. Su CJ, Tryon VV, Baseman JB. Cloning and sequence analysis of cytadhesin PI gene from Mycoplasma pneumoniae. Infect Immun 1987; 55(12):3023-3029.
14. Campbell LA, Perez MM, Hamilton DJ, Kuo CC, Grayston JT. Detection of Chlamydia pneumoniae by polymerase chain reaction. J Clin Microbiol 1992; 30(2):434-439.
15. Fine MJ, Singer DE, Hanusa BH, Lave JR, Kapoor WN. Validation of a pneumonia prognostic index using the MedisGroups Comparative Hospital Database. Am J Med 1993; 94(2):153—159.
16. Niederman MS, Bass JB Jr, Campbell GD, et al. Guidelines for the initial management of adults with community-acquired pneumonia: diagnosis, assessment of severity, and initial antimicrobial therapy. American Thoracic Society. Medical Section of the American Lung Association. Am Rev Respir Dis 1993; 148(5):1418-1426.
17. File TM Jr, Segreti J, Dunbar L, et al. A multicenter, randomized study comparing the efficacy and safety of intravenous and/or oral levofloxacin versus ceftriaxone and/or cefuroxime axetil in treatment of adults with community-acquired pneumonia. Antimicrob Agents Chemother 1997; 41(9):1965-1972.
18. Woodhead MA, Macfarlane JT, McCracken JS, Rose DH, Finch RG. Prospective study of the aetiology and outcome of pneumonia in the community. Lancet 1987; 1 (8534):671 -674.
19. Klimek JJ, Ajemian E, Fontecchio S, Gracewski J, Klemas B, Jimenez L. Community-acquired bacterial pneumonia requiring admission to hospital. Am J Infect Control 1983; 11(3):79—82.
20. Halm EA, Fine MJ, Marrie TJ, et al. Time to clinical stability in patients hospitalized with community-acquired pneumonia: implications for practice guidelines [see comments], JAMA 1998; 279(18):1452-1457.
21. Plouffe JF, Herbert MT, File TM Jr, et al. Ofloxacin versus standard therapy in treatment of community-acquired pneumonia requiring hospitalization. Pneumonia Study Group. Antimicrob Agents Chemother 1996; 40(5):1175-1179.
22. Plouffe JF, Breiman RF, Facklam RR. Bacteremia with Streptococcus pneumoniae. Implications for therapy and prevention. Franklin County Pneumonia Study Group. JAMA 1996; 275(3) :194-198.
23. Lonks JR, Medeiros A A. High rate of erythromycin and clarithromycin resistance among Streptococcus pneumoniae isolates from blood cultures from Providence, R.I. Antimicrob Agents Chemother 1993; 37(9):1742-1745.
24. Mandell, GL, Donowitz GR. Acute pneumonia. In: Mandell GL, Bennet JE, Dulin R, eds. Principles and practice of infections diseases. Philadelphia: Churchill Livingstone, 2000: 717-743.
