Levofloxacin

Study objectives: To evaluate the efficacy and tolerability of levofloxacin (L) as monotherapy in patients with severe community-acquired pneumonia (CAP) in comparison with therapy using a combination of cefotaxime (C) plus ofloxaein (O).

Design: Prospective, randomized 1:1, comparative, open, parallel-group study.

Setting: Multinational study with 149 sites.

Patients: A total of 398 randomized patients who had been admitted to the ICU with severe CAP without shock, including 308 patients in a modified intent-to-treat population and 271 patients in the per-protocol (PP) population (L group, 139 patients; C + O group, 132 patients).

Interventions: Therapy with levofloxacin (500 mg IV, q12h) vs therapy with a C + O combination (C, 1g IV, q8h; O, 200 mg IV, q12h) for 10 to 14 days.

Measurements and results: The main end point was the clinical efficacy at the end of treatment (ie, the test-of-cure [TOC] visit). The statistical hypothesis was the noninferiority of L therapy to C + O therapy with a 2.5% [alpha] risk (unilateral) and a 15% maximum set difference. At the TOC visit, a clinical success was observed in 79.1% of patients (L group) and 79.5% of patients (C + O group) in the PP population (difference, -0.4%; 95% confidence interval [CI], -10.79 to 9.97% without adjustment for simplified acute physiology score [SAPS] II at inclusion; difference, -0.3%; 95% CI, -10.13 to 9.58% with adjustment for SAPS II). A satisfactory bacteriologic response was present in 73.7% of L group patients and 77.5% of C + O group patients, including responses of 75.7% and 70.3%, respectively, in the L group and C + O group in the Streptococcus pneumoniae-documented population. In the safety analysis, 20 patients in the L group (10.3%) and 16 patients in the C + O group (8.0%) experienced at least one adverse event that was considered to be treatment-related.

Conclusion: L therapy was at least as effective as the combination therapy of C + O in the treatment of a subset of patients with CAP requiring ICU admission. This conclusion cannot be extrapolated to patients requiring mechanical ventilation or vasopressors (ie, those patients in shock).

Key words: community-acquired pneumonia; ICU; levofloxacin; Streptococcus pneumoniae

Abbreviations: ATS = American Thoracic Society; C = cefotaxime; CAP = community-acquired pneumonia; FI[O.sub.2] = fraction of inspired oxygen; L = levofloxacin; mITT = modified intention-to-treat; O = ofloxacin; PP = per protocol; SAPS = simplified acute physiology score; TOC = test-of-cure.

**********

In patients with community-acquired pneumonia (CAP) requiring intensive care, the most common causative organisms are Streptococcus pneumoniae, Legionella pneumophila, and Haemophilus influenzae. (1) In addition, some series have reported (1) that Staphylococcus aureus could be a frequent pathogen, that Gram-negative aerobic organisms are identified with increasing frequency and, finally, that Chlamydia pneumoniae and Mycoplasma pneumoniae could be associated with severe CAP.

Levofloxacin (L) is a fluoroquinolone that is active against most of the pathogens mentioned, especially S pneumoniae, including strains with decreased penicillin susceptibility. (2-6) This antibiotic is available in an IV and an oral presentation, the latter exhibiting an excellent, near-complete bioavailability, allowing an IV-to-oral administration switch as soon as oral feeding is resumed. (7-10) The dose used in Europe is 500 mg qd or bid, according to the severity of CAP.

Its clinical activity in pneumococcal CAP has been well-documented in various clinical trials in Europe and the United States. (8-12) However, although at least four studies (8-10,12) have included patients with severe infections, most patients included in those studies usually presented with mild-to-moderately severe CAP.

Recommendations for the empiric treatment of patients with CAP have been announced and updated by numerous scientific groups. (1,13-18) In the case of severe CAP requiring ICU admission, treatment combining a [beta]-lactam (amoxicillin with or without clavulanic acid, cefotaxime (C), and ceftriaxone) and a macrolide or a fluoroquinolone is usually preferred. The limited data supporting monotherapy with macrolides or fluoroquinolones for patients with severe pneumococcal CAP explain the recommendation of combination treatment for all patients hospitalized in the ICU. (14,18)

The lack of studies in patients with severe CAP who have been admitted to an ICU comparing an antimicrobial combination therapy with monotherapy exhibiting an equivalent antibacterial spectrum and the microbiological and pharmacokinetic properties of L provided a strong rationale for performing a study comparing L with a recommended antimicrobial combination therapy such as C plus ofloxacin (O). The trial was conducted between January 2000 and November 2002 at 149 institutions in three countries.

MATERIALS AND METHODS

Patient Selection

Patients were eligible for enrollment in the study if they were adult (ie, age > 18 years) and had severe CAP requiring ICU admission. CAP was defined by the presence of a new radio graphic pulmonary infiltrate seen at the initial presentation or occurring within 48 h following hospitalization, and associated with a total leukocyte count of > 10,000 cells/[micro]L or < 4,500 cells/[micro]L and fever (oral or axillary or inguinal temperature of > 38[degrees]C, or rectal or aural temperature of > 38.5[degrees]C), plus at least one of the following clinical signs: cough of recent onset or recently exacerbated; purulent sputum of recent appearance; dyspnea., chest pain; crackling rales; and/or signs of consolidation on pulmonary auscultation. The severity of CAP, justifying admission to the ICU, was confirmed by the presence either of a major criterion or two minor criteria. A major criterion was a Pa[O.sub.2]/fraction of inspired oxygen (FI[O.sub.2]) ratio of <250 mm Hg requiring invasive or noninvasive ventilation. Minor criteria were a respiratory rate of > 30 breaths/rain, Pa[O.sub.2] of <60 mm Hg, or PaC[O.sub.2] of >50 mm Hg at an FI[O.sub.2] of 0.21, a chest radiographic involvement of more than a single lobe, and altered mental status.

Patients hospitalized during the previous month, who were admitted from a nursing home, who developed pneumonia > 48 h after hospital admission, or who had previously received antibiotic therapy for this CAP episode could not be enrolled in the study. However, outpatients who had been treated for > 48 h with antibiotics and had been admitted to the ICU due to lack of response to treatment could be included in the study unless the prior antibiotic they had received was L, sparfloxacin, or ceftriaxone combined with O. Exclusion criteria were the presence of a CAP-causative pathogen known to be resistant to the antibiotics used in the study, infectious disease requiring concomitant antimicrobial treatment, septic shock (as defined by the criteria in the study by Bone et al (19) prior to study inclusion, life expectancy of < 2 days, underlying terminal malignancy, cystic fibrosis, or suspected active tuberculosis, CD4 cell count of < 50 cells/[micro]L secondary to HIV infection, immunosuppression (ie, leukocyte count, < 1,000 cells/[micro]L or ongoing radiation treatment), hypersensitivity, or contraindications to any study medication. Patients who were unlikely to comply with the protocol requirements, having participated in another study or having taken another investigational drug in the month prior to study inclusion, or those not meeting the legal requirements for participation in an investigational study were also excluded. The study protocol was approved by the ethics committees in the respective countries, and written consent was obtained from all patients or their relatives before study enrollment. Patients were recruited by 87 centers in three countries (France, Tunisia, and South Africa).

Study Design

This prospective, multicenter, multinational study used a randomized, open, parallel-group design to compare therapy with IV L with combination therapy using IV C with IV O. Randomization was performed locally on the basis of a centrally established code. Patients were to be treated for a period ranging from 10 to 14 days. However, the duration was extended up to 21 days in cases of CAP due to Legionella sp or associated with purulent pleurisy. For L and O, a switch to oral medication was allowed once the oral route of administration was deemed possible. Study drug dosage adjustments, which are required for patients with renal impairment, were based on the estimated creatinine clearance value.

Patients in the L group were initially administered 500 mg L by IV infusion over 60 min bid. Thereafter, L could be given as a 500-rag tablet bid. Patients receiving the comparator regimen were given 1g C by IV infusion over 20 to 60 min tid and 200 mg O by IV infusion over 60 min bid. Oral O was administered as a 200-mg tablet bid. A switch to monotherapy was possible when the identified causal organism was either S pneumoniae (0 therapy could be stopped) or Legionella sp (C therapy could be stopped).

Patient evaluations were performed before treatment, 3 days after the beginning of treatment, 1 day after the end of therapy (the test-of-cure [TOC] visit), and 21 to 45 days after the end of study medication (follow-up visit). On study inclusion, the following characteristics were collected: age; gender; weight; underlying clinical conditions; the presence of a prior antimicrobial failure for this CAP episode; duration of symptoms prior to study enrollment; severity of illness; vital sign abnormalities; number of lobes involved on a chest radiograph; need for mechanical ventilation; and the presence of an ARDS. Severity of illness was assessed by simplified acute physiology score (SAPS) II. (20) ARDS was defined by the usual criteria. (21) Biochemical and hematologic tests were performed and bacteriologic samples were obtained according to the investigator's need or the center's practice, but the measurement of arterial blood gases was compulsory on study enrollment. Establishment of the etiologic diagnosis required the isolation of bacteria in significant quantities from a sample of lower respiratory tract secretions (endotracheal aspiration sample, > [10.sup.6] cfu/mL; protected brush catheter sample, > 10:3 cfu/mL; and BAL fluid, > [10.sup.4] cfu/mL), the isolation of a definitive pathogen from a blood or pleural fluid culture, the detection of L pneumophila serogroup 1 antigen in urine, or a significant antibody titer rise in paired serologic testing for L pneumophila, C pneumoniae, and M pneumoniae. A chest radiograph and a complete physical examination were repeated at control visits performed during hospitalization. Bacteriologic and biological sampling were left to the discretion of the investigator. Finally, the follow-up visit, performed 21 to 45 days after the end of treatment, could be performed either by the investigator or by another physician.

Patient Evaluation

The primary end point was the assessment of clinical efficacy at the TOC visit. The clinical response, assessed by the investigator, was based on the outcome of the signs and symptoms collected at study inclusion, on changes in the chest radiograph findings, and, when appropriate, on bacteriologic data. A cure was defined as the disappearance of acute signs and symptoms, and the improvement of radiographic abnormalities, both related to CAP, with no requirement for further antimicrobial therapy. Failure was defined by any modification of antimicrobial treatment that was motivated by the isolation of a causative organism resistant to the study medication, lack of clinical improvement 3 days after treatment initiation (ie, worsening or persistent fever or hypothermia, or worsening of pulmonary infiltrates or respiratory function as assessed by Pa[O.sub.2]/FI[O.sub.2] ratio), insufficient clinical response, secondary clinical deterioration due to sepsis-related complications (ie, septic shock or ARDS), or lower respiratory tract superinfection, and the occurrence of an adverse event. Finally, if a patient died during the first 48 h of therapy or developed an infection outside the lower respiratory tract requiring a modification of antimicrobial therapy, the outcome was considered to be indeterminate but was analyzed as a failure.

Secondary efficacy parameters included the bacteriologic response at the TOC visit, clinical and bacteriologic responses at the follow-up visit, overall and 28-day mortality rates, and length of stay in the ICU. Patients who were receiving mechanical ventilation at study inclusion were included in a subgroup in which clinical efficacy at the TOC visit, and the overall and 28-day mortality rates were assessed. Outcomes at the follow-up visit were defined by the criteria used at the TOC visit. Moreover, a failure at the TOC visit or the occurrence of death or a new episode of pneumonia between the TOC and the follow-up visits also were considered to be failures at the follow-up visit. When a patient was lost to follow-up and when subsequent antimicrobial therapy was required for reasons other than lower respiratory tract infection, the outcome was considered to be indeterminate but was analyzed as a failure. In bacteriologically documented patients, bacteriologic outcome at the TOC visit was considered to be satisfactory if control cultures were negative (eradication), were not available (presumed eradication), or were positive with identification of pathogens that were different from the causative pathogens (colonization), but were associated in all cases with a clinical cure. Finally, when a clinical failure was associated with the noneradication of the initial pathogen (persistence) or with the isolation of pathogens that were different from the baseline pathogens (superinfection), or when there were no samples for culture (presumed persistence), the bacteriologic outcome was defined as unsatisfactory. At the follow-up visit, a similar evaluation was performed. Moreover, when relapse or reinfection occurred between the TOC and follow-up visits, the outcome was considered to be a failure.

The study was also designed to analyze safety. Adverse events, which were recorded at every visit, were categorized and summarized according to the Coding Symbols for a Thesaurus of Adverse Reaction Terms system. (22) The events were assessed by the investigator for severity and relation to the study medication.

Statistical Analysis

The noninferiority of L would be demonstrated if the two-tailed 95% confidence interval for the difference in clinical success at the end of treatment with the study medication (TOC visit) between the two groups (ie, the response in the L group minus the response in the C + O group) included 0 and the lower boundary was not [less than or equal to] 15%. The main analysis was performed with and without adjustment for SAPS II at study inclusion. Assuming an estimated 70% success rate, a maximum 15% difference between groups, an a risk set at 0.05, and a power of 80%, a sample size of 124 evaluable patients per study group would be sufficient. The principal efficacy analysis was performed with the same hypotheses except that a bilateral 5% approach (2.5% one-sided approach) was used instead to follow the present European recommendations. (23,24)

The two analyzed populations were the modified intention-to-treat (mITT) population and the per-protocol (PP) population [the primary analysis]. The mITT population included all patients with a confirmed diagnosis of CAP who had completed the inclusion procedures, had received the study medication for at least 2 days, and had at least one available clinical evaluation. The criteria for inclusion in the PP population were inclusion in the mITT analysis and no major protocol violation, receipt of the study medications for at least 7 days (2 days in the case of a clinical failure), and an appropriate evaluation of efficacy. Only patients included in the PP population in whom a causative pathogen for CAP was isolated on study inclusion were considered to be bacteriologically evaluable. The valid-for-safety population included all randomized patients who had taken at least one dose of the study medication.

Patient characteristics were described by frequencies for categoric data, and by mean, SD, and median for numeric variables in both treatment groups. Categoric variables were compared using the [chi square] test, or the Fisher exact test when the [chi square] test was not appropriate. Continuous variables were compared using the Student t test. Differences were considered to be significant for variables yielding a p value of [less than or equal to] 0.05. To assess that the severity of CAP was similar in the two groups, we compared SAPS II and the usual factors of a complicated outcome. Finally, in the PP population, we identified patients with pneumococcal CAP and/or bacteremic CAP, and we compared the clinical and bacteriologic efficacy of both study medications.

The study design was developed jointly by the study sponsor and an independent scientific study committee. Statistical analysis was performed using a statistical software package (Effi-Stat; Paris, France).

RESULTS

Baseline Characteristics of the Patients Over a period of 34 months, 398 patients were enrolled in the study, 308 of whom were included in the mITT analysis. Of these, 149 patients were treated with L and 159 with C + O. The main reasons for exclusion from mITT analysis were a CAP misdiagnosis (n = 62) and a duration of therapy of < 2 days (n = 25). From the mITT population, 271 patients were included in the PP population. The trial population profile is depicted in Figure 1.

[FIGURE 1 OMITTED]

The baseline characteristics of patients included in the mITT and PP analyses are summarized in Table 1. There were no differences between the two groups in the distribution of patients according to demographic characteristics, comorbid conditions, and severity of CAP episode. In the mITT populations, 65 patients in the L group (43.6%) and 73 patients in the C + O group (45.9%) exhibited a Pa[O.sub.2]/FI[O.sub.2] ratio of < 250 mm Hg, which indicated the need for mechanical ventilation. For the remaining patients, in the L group (n = 84) 62 patients (73.8%) exhibited a multilobar involvement, 33 patients (39.3%) exhibited a Pa[O.sub.2]/FI[O.sub.2] ratio of < 250 mm Hg, and 21 patients (25%) exhibited a multilobar involvement that was associated with a Pa[O.sub.2]/FI[O.sub.2] ratio of < 250 mm Hg. In the C + O group (n = 86), 65 patients (75.6%), 39 patients (45.3%), and 30 patients (34.7%), respectively, exhibited these abnormalities.

Causative organisms were identified in 168 patients in the mITT analysis (L group, 82 patients; C + O group, 86 patients) and in 147 patients in the PP analysis (L group, 76 patients; C + O group, 71 patients). Main organisms were S pneumoniae and H influenzae (Table 2). Combined pathogens were present in 12 patients (L group) and 17 patients (C + O group), respectively. The most frequent combination was S pneumoniae plus H influenzae. Atypical pathogens were incriminated in only three cases of combined etiologies (all concerned a C pneumoniae infection). Legionella sp was always identified as a single pathogen. Blood culture results were positive in 39 patients (12.7%) overall (L group, 20 patients [13.4%]; and C + O group, 19 patients [11.9%]). S pneumoniae was isolated in blood cultures in 10 patients in the L group and in 15 patients in the C + O group.

Duration of Treatment

The duration of treatment was similar in the two groups (p = 0.13). In the mITT population, the mean ([+ or -] SD) duration of treatment was 11.1 [+ or -] 4.6 days in the L group and 11.5 [+ or -] 3.4 days in the C + O group. A switch to oral L was performed 6.9 [+ or -] 2.9 days after the initiation of treatment for 42 patients (28%) in the L group. In the C + O group, a switch to a monotherapy based on C (n = 30) or O (n = 23) occurred in 53 patients (33%) 6.4 [+ or -] 2.9 days after initiation of treatment. The mean duration of the ICU stay was 11.9 [+ or -] 9.1 days (for survivors, 11.9 [+ or -] 9.4 days) in the L group and 12.9 [+ or -] 10.5 days (for survivors, 12.0 [+ or -] 9.7 days) in the C + O group.

Efficacy

Table 3 presents the results of the analysis of clinical efficacy at the TOC and follow-up visits. In the overall population, the results are consistent with a noninferiority of L to C + O since the lower bounds of the confidence intervals always exceed -15%. The main causes of failure, at the TOC visit in the PP population, are presented in Table 4.

In patients receiving mechanical ventilation, the clinical cure rates at the TOC visit were as follows: PP analysis: L group, 63.0% (46 of 73 patients); C + O group, 71.8% (51 of 71 patients; p = 0.26); mITT analysis: L group, 60.5% (46 of 76 patients); C + O group, 70.7% (58 of 82 patients; p = 0.18).

In the bacteriologically evaluable patients, the satisfactory bacteriologic responses at the TOC and follow-up visits were similar in both treatment groups (Table 5). Most of the nonsatisfactory responses were due to a presumed persistence or a superinfection. In the L group, there were four cases of nosocomial pneumonia, for which the causative organisms were methicillin-resistant S aureus (n = 2) and Pseudomonas aeruginosa (n = 2). In the C + O group, six patients exhibited nosocomial pneumonia due to methicillin-resistant S aureus (n = 2) and P aeruginosa (n = 4).

Mortality

According to the SAPS II values at study inclusion, the expected mortality rate in the mITT population was 20.5%. The overall observed mortality rate in this population was 14.3%. In the two treatment groups, the 28-day and overall mortality rates were similar, whichever population was studied (Table 6).

Safety Results

Overall, one or more adverse events were recorded in 138 patients in the L group and in 141 patients in the C + O group (p = 0.83). Most events were unrelated to the study medication. Adverse events requiring study drug discontinuation and drug-related events were infrequent (Table 7). In the L group, the adverse events requiring the discontinuation of treatment were cytolytic liver injury (n = 1), allergic rash (n = 1), leukopenia (n = 1), tendon rupture (n = 1), and agitation and persecutory delusion (n = 1). In the C + O group, the adverse events requiring the discontinuation of treatment were allergic rash (n = 3) and thrombocytopenia (n = 1).

DISCUSSION

Our trial was designed to compare the efficacy of L with that of C + O in patients exhibiting severe CAP requiring ICU admission. In two groups that were comparable in all respects, we demonstrated that, irrespective of the end point used, L therapy was not inferior to the C + O combination therapy in terms of clinical success, bacteriologic eradication, safety, and mortality.

In patients with severe CAP, S pneumoniae, Legionella sp, H influenzae, and S aureus were the primary causative pathogens. (1,25-28) The relevance of Gram-negative bacilli, such as the Enterobacteriaceae and P aeruginosa, also appears to be significant, if not increasing. Finally, cases of severe CAP due to C pneumoniae or M pneumoniae have been reported. (25,28) To avoid inadequate and/or ineffective antibiotic therapy, adversely affecting the prognosis, all experts have proposed (25,26) the institution of an empiric treatment covering all potential pathogens. Thus, the North American recommendations (1,14,15,18) state that, in all patients in whom a P aeruginosa infection is not suspected, the use of an extended-spectrum cephalosporin (ie, C or ceftriaxone) or a [beta]-lactam-[beta]-lactamase inhibitor in combination with either an IV macrolide or an IV respiratory fluoroquinolone is recommended. The antibacterial and pharmacokinetic properties of L provided a good rationale for its evaluation in patients with mild-to-moderately severe CAP. In the studies performed, (8,9,11,12) L appeared to be at least as effective as comparators. Thus, it was appropriate to propose its use in monotherapy for the treatment of CAP in inpatients hospitalized on a medical ward. (1,14,15,18) This approach was validated by a metaanalysis performed by Salkind et al (29) on 13 studies comparing a respiratory quinoloue agent with an oral macrolide agent or a [beta]-lactam. This analysis showed a statistically significant advantage in favor of the fluoroquinolone. However, although the etiologic pattern is often similar in both ICU and non-ICU patients, a respiratory quinolone has not yet been recommended as monotherapy for patients admitted to the ICU for severe CAP. This point was again stressed by the Community-Acquired Pneumonia Committee from the Infectious Diseases Society of America (18) and is due to the lack of efficacy data on fluoroquinolones as monotherapy for the treatment of severe CAP. Our work provides an interesting contribution to the therapeutic approach in these patients. In a population with > 250 evaluable patients, therapy with L was not found to be inferior to therapy with C + O. Furthermore, L and the C + O combination appear to have similar efficacy in patients with severe pneumococcal CAP or severe pneumococcal CAP with bacteremia. Fogarty et al (30) recently reported a trial including 269 seriously ill patients with CAP that was designed to compare L (500 mg IV, q24h) with ceftriaxone sodium (1 to 2 g, q24h) plus erythromycin (500 to 1,000 mg IV, q6h). With a clinical efficacy close to 85% in both groups, L monotherapy appeared to be as effective as the combination therapy. Although some differences existed between the study populations (20.6% patients required mechanical ventilation in the study by Fogarty et al (30) vs 51.3% in ours) and between the study regimens (ie, L regimen in the study by Fogarty et al, (30) 500 mg per 24 h; our study, 500 mg q12h), these trials demonstrate the efficacy of L used as monotherapy in patients with severe CAP.

To be included in our study, patients had to present with severe CAP to justify ICU admission. Although there is no universally accepted definition of severe CAP, ICU admission could be justified by criteria such as those originally recommended by, or revised from, the American Thoracic Society (ATS) and the European Study on CAP Committee. (1,13,31) The criteria for enrollment in our study were chosen from among these criteria. Ewig et al (32) have demonstrated that the original ATS criteria were indeed sensitive but were poorly specific for the severity of CAP. They proposed reducing the number of criteria and combining them under major criteria (eg, shock or mechanical ventilation) and minor criteria (eg, multilobar involvement, Pa[O.sub.2]/FI[O.sub.2] ratio of < 250 mm Hg, and systolic BP of < 90 mm Hg). Therefore, they showed that CAP could be considered as severe when a patient had one major criterion or two minor criteria. This work was taken into account by the ATS in 2001. (1) Obviously, our study, having been designed in 1999, could not use these revised criteria as inclusion criteria. Similarly, we could not use the British Thoracic Society rules that were published in 2001. (17) However, it must be stressed that, in each study group, about 45% of patients had severe pneumonia, as determined by a low Pa[O.sub.2]/FI[O.sub.2] ratio that indicated the need for mechanical ventilation. In the remaining patients, according to the criteria proposed by Ewig et al, (32) 30% of patients exhibited a multilobar involvement associated with a Pa[O.sub.2]/ FI[O.sub.2] ratio of < 250 mm Hg. However, according to the criteria proposed by the British Thoracic Society, we must emphasize that about 85% of patients without a Pa[O.sub.2]/FI[O.sub.2] ratio of < 250 mm Hg who required mechanical ventilation exhibited either two core adverse prognostic features (eg, confusion and respiratory rate of [greater than or equal to] 30 breaths/min) or one criterion associated with one or two additional adverse prognostic features (eg, multilobar involvement or Pa[O.sub.2] of < 60 mm Hg). The SAPS II values calculated at study inclusion are another confirmation of the severity level of the population, since the SAPS II-predicted mortality rate was around 20%. The overall and 28-day mortality rates were, in the valid-for-safety population, 17.5% and 15.5%, respectively, in the L group, and 22.4% and 18.4%, respectively, in the C + O group. These results were in accordance with literature data reporting that the mortality rate for severe CAP ranges from 15 to 30%. (25,26,32-36)

The most frequent causative organisms recovered in our study were S pneumoniae and H influenzae. Although patients with chronic respiratory insufficiency, those who have recently received antibiotic therapy, or those who have had a recent hospital stay accounted for a significant number of the patients in our series, no patient presented with an infection related to P aeruginosa. These results are in disagreement with those originating in the United States but are consistent with data from France. (25,35,37) In the bacteriologically evaluable patients, about 50% of patients presented with pneumococcal CAP. In these patients, the bacteriologic and clinical success rates were 75.7% in the L therapy group and 70.3% in the combination therapy group. Bacteremia was present in 20 to 25% of patients with bacteriologically documented CAP. The most frequent pathogen was S pneumoniae. As for the entire bacteriologically evaluable population, the success rates were similar in both groups. They were 68.4% (bacteremic CAP) and 70.0% (bacteremic pneumococcal CAP) in the L group, and 62.5% (bacteremic CAP) and 53.8% (bacteremic pneumococcal CAP) in the comparator group. Although the number of patients was low, our data confirm the efficacy of L used as a single antibiotic in the treatment of severe pneumococcal CAP, bacteremic or otherwise, and are in agreement with a recent publication reported by Kahn et al. (38) Nine phase III and IV trials evaluating the efficacy of L in patients with CAP were reviewed. They included 497 evaluable patients with pneumococcal CAP. Among them, 108 patients had CAP-associated pneumococcal bacteremia. Success occurred in 95% of patients (472 of 495 patients) and 90.7% of patients (98 of 108 patients), respectively. Until the last few years, it was always assumed that antibiotic monotherapy was sufficient to treat patients with pneumococcal CAP. However, three publications (39-41) have suggested that, in the case of pneumococcal CAP with bacteremia, a dual combination therapy including a macrolide reduces mortality. Although these studies are retrospective, nonrandomized, and not designed to provide this type of evaluation, the results are intriguing. A tentative explanation could be either additive activity against S pneumoniae, the concurrent presence of atypical pathogens, or the immunomodulating effect of macrolides. In our population, no immunomodulating effect or broadening of the antibacterial spectrum with the combination therapy was to be expected. No apparent additive effect of O and C was observed, since no difference in mortality between monotherapy and combination therapy was evidenced. However, the antipneumococcal activity of O is low, and it might be interesting in the future to compare the efficacy of L administered alone with that of combination therapy with a [beta]-lactam active against S pneumoniae.

Our study has the following limits. First it was not performed under double-blind conditions. This was due to the practical issue of implementing the double-blinded administration of study medication. Second, the bacterial documentation was not to be based on specific sampling techniques, and there was no centralization of bacterial samples. Our study was primarily designed as a clinical evaluation, and was not intended to provide bacterial epidemiology data on respiratory pathogens and their antibiotic susceptibility. Third, we did not calculate the Fine score of our patients. The decision was based on the fact that this score was not established for the severity assessment within the group of patients with severe CAP but, rather, was established for the assessment within the overall population of CAP patients. In consequence, we used SAPS II, which is a severity score that is widely used in intensive care practice. Fourth, evidence of shock precluded the inclusion of patients in the study. As it had been demonstrated (25,26,33,37) that shock was associated with a high mortality rate and an early occurrence of death, we believed that the inclusion of such patients could have led to a too important increase in the recruitment necessary to provide the number of evaluable patients for a valid interpretation. Finally, although, in mechanically ventilated patients, the cure rate at the TOC visit, and the 28-day and overall mortality rates were not statistically different in both treatment groups according to univariate analysis, we were unable to demonstrate in these patients the noninferiority of L to C + O. The low number of mechanically ventilated patients included in our study probably explains our inability to demonstrate this noninferiority.

In conclusion, L therapy (500 mg q12h) was at least as effective as C + O combination therapy in the treatment of a subset of patients with CAP requiring ICU admission. This conclusion cannot be extrapolated to patients requiring mechanical ventilation or vasopressors (for treatment of shock). Although this latter point represents a major limitation of our study, we believe it would be interesting to perform studies comparing monotherapy with L and combination therapy using an extended-spectrum cephalosporin or a [beta]-lactam-[beta]-lactamase inhibitor with either a macrolide or even L itself, and including a sufficient number of patients to definitively assess the value of monotherapy in the patients with the most severe disease (ie, those with septic shock and those requiring mechanical ventilation).

APPENDIX: LEVOFLOXACIN STUDY GROUP

France

Herve Aube, MD, General Hospital, Dijon; Jean-Pierre Auffray, MD, Sainte Marguerite Hospital, Marseille; Olivier Barnoud, MD, University Hospital, Grenoble; Gilles Beaucaire, MD, General Hospital, Tourcoing; Patrice Beck, MD, Hospital Center, Saint Lo; Bernard Bedock, MD, Hospital Center, Annonay; Jean-Pierre Bedos, MD, Andre Mignot Hospital, Le Chesnay; Michel Bemer, MD, Bel Air Hospital, Thionville; Christian Bengler, MD, Hospital Center, Nimes; Claude Beuscart, MD, Hospital Center, Saint Brieuc; Edouard Bollaert, General Hospital, Nancy; Thierry Boulain, MD, La Source Hospital, Orleans; Alain Bozier, MD, Hospital Center, Avranches; Francois Brivet, MD, Antoine Beclere Hospital, Clamart; Daniel Caen, MD, Gilles de Corbeil Hospital, Corbeil-Essonnes; Gilles Capellier, MD, Jean Minjoz Hospital, Besancon; Vincent Castelain, MD, University Hospital, Le Kremlin Bicetre; Jean-Luc Chagnon, MD, Hospital Center, Valenciennes; Vonick Chenu, MD, Hospital Center, Dax; Pascal Chevallier, MD, Hospital Center, Brive; Bernard Clair, MD, Raymond Poincare Hospital, Garches; Marc Clavel, MD, Hospital Center, Carcassonne; Yves Cohen, MD, Avicenne Hospital, Bobigny; Christian Combe, MD, ViIlefranche Regional Hospital, Gleize; Jean-Michel Coudray, MD, Hospital Center, Longjumeau; Jean-Michel Couland, MD, Hospital Center, Montfermeil; Pierre Courant, MD, Hospital Center, Avignon; Patrick Courtin, MD, Hospital Center, Martigues; Serge Delayance, MD, Clinique des Cedres, Cornebarrieu; Patrick Dolan, MD, Hospital Center, Laval; Yves Domart, MD, Hospital Center, Compibgne; Rene Dome, MD, St Luc-St Joseph Hospital, Lyon; Didier Dubois, MD, Hospital Center, Arras; Jean-Pierre Dubois, MD, Hospital Center, Villiers St Denis; Patricia Dulbecco, MD, Hospital Center, Antibes-Juan Les Pins; Andre Dupont, MD, Hospital Center, Fontainebleau; Marc Dupont, MD, St Joseph Hospital, Marseille; Alain Durocher, MD, Regional University Hospital, Lille; Richard Faitg, MD, Hospital Center, Thonon Les Bains; Christian Floriot, MD, Paul Moral Hospital, Vesoul; Francois Fraisse, MD, Delafontaine Hospital, St Denis; Annie Freche, MD, Hospital des Broussailles, Cannes; Claude Gabinsky, MD, Saint Andre Hospital, Bordeaux; Jean-Lou Galiacy, MD, Saint Esprit Hospital, Agen; Pierre Gareia, MD, Saint Roch Hospital, Nice; Bernard Garrigues, MD, Hospital Center, Aix En Provence; Bernard Gauche, MD, Robert Boulin Hospital, Libourne; Pierre Geffe, MD, Belle Isle Hospital, Metz; Bernard Georges, MD, Rangueil Hospital, Toulouse; Main Gerard, MD, Brabois Hospital, Vandoeuvre les Nancy; Michel Gonzales, MD, Robert Bisson Hospital, Lisieux; Michel Grinand, MD, Moulin du Pre Hospital, St Nazaire; Dominique Guelon, MD, Gabriel Montpied Hospital, Clermont Ferrand; Anne-Marie Guerin-Robardey, MD, Hospital Center, Beauvais; Moldi Hamrouni, MD, Fontenoy Hospital, Le Coudray; Marie-Helene Hausermann, MD, Henri Mondor Hospital, Aurillac; Gilles Hilbert, MD, Pellegrin Hospital, Bordeaux; Michel Hira, MD, Hospital Center, Chateauroux; Bernadette Hugot, MD, Purpan Hospital, Toulouse; Francoise Jungfer, MD, Hospital Center, Sens; Michel Kaidomar, MD, Hospital Center, Frejus; Jean-Michel Korach, MD, Hospital Center, Chalons en Champagne; Jacques Labrousse, MD, European Hospital Georges Pompidou, Paris; Guillaume Lalande, MD, Hospital des Hauts Clos, Troyes; Gilbert Laplatte, MD, Louis Pasteur Hospital, Colmar; Jean Roger Le Gall, MD, Saint Louis Hospital, Paris; Le Quoc Viet, MD, Morey Hospital, Chalon sur Saone; Christian Lemaire, MD, Victor Provost Hospital, Roubaix; Alain Leon, MD, Hospital Center, Reims; Jacques Leroy, MD, Charles Nicolle Hospital, Rouen; Jean-Francois Loriferne, MD, Sainte Camille Hospital, Bry sur Marne; Philippe Lutun, MD, Hautepierre Hospital, Strasbourg; Jacques Maindivide, MD, Hospital Center, Saintes; Laurent Mandin, MD, Hospital Center, Gap; Bernard Mankikian, MD, Clinique Saint Gatien, Tours; Olivier Marie, MD, Saint-Louis Hospital, Paris; Claude Martin, MD, Nord Hospital, Marseille; Bernard Melon, MD, Francois Mitterrand Hospital, Pan; Asri Merouani, MD, Hospital Center, Alencon; Gerard Meyer, MD, General Hospital Center, Senlis; Olivier Millet, MD, Hospital Center, Lons le Sannier; Benoit Misset, MD, Saint Joseph Hospital, Paris; Philippe Morel, MD, Hospital Center, Cambrai; Christian Moret, MD, Hospital Center, Gueret; Michael Moriconi, MD, Hospital de Cornouailles, Quimper; Serge Moulront, MD, Hospital Center, Dunkerque; Francois Nicolas, MD, General Hospital, Moulins; Jean-Philippe Nouveau, MD, Hospital Center, Le Havre; Edouard Obadia, MD, Claude Galien Hospital, Quincy sous Senart; Christophe Obez, MD, Clinique des Sources, Nice; Georges Offenstadt, MD, Saint Antoine Hospital, Paris; Jean-Rene Ordonneau, MD, Regional University Hospital Laennec, Nantes; Bernard Page, MD, Ambroise Pare Hospital, Boulogne; Yves Page, MD, University Hospital, Saint-Etienne; Jean-Louis Pallot, MD, Andre Gregoire Hospital, Montreuil; Bruno Palmier, MD, Hospital d'Instruction des Armees Sainte Anne, Toulon; Yves Parer Aubas, MD, Lapeyronnie Hospital, Montpellier; Antoine Parrot, MD, Tenon Hospital, Paris; Rene-Gilles Patrigeon, MD, Desgenettes Hospital, Lyon; Herve Peneau, MD, Hospital Center, Calais; Dominique Perrotin, MD, Bretonneau Hospital, Tours; Paul Petit, MD, Edouard Herriot Hospital, Lyon; Michel Pinsard, MD, Hospital Center, Dieppe; Patrick Pinta, MD, La Gespe Hospital Center, Tarbes; Raphael Pitti, MD, Military Hospital Legouest, Metz; Geoffroy Place, MD, Max Fourestier Hospital, Nanterre; Jean-Francois Poussel, MD, Bon Secours Hospital, Metz; Antoine Rabbat, MD, Hotel Dieu Hospital, Paris; Jean-Philippe Rigaud, MD, Hospital Center, Le Havre; Alain Rime, MD, Clinique du Parc, Castelnau le Lez; Marie-Aline Robaux, MD, Hospital de la Cavale Blanche, Brest; Rene Robert, MD, University Hospital, Poitiers; Jean-Marc Rodolfo, MD, Hospital Center, Auch; Alain Rosant, MD, Hospital Center, Mont Saint Martin; Christian Roth, MD, Hospital Center, Annemasse; Olivier Ruyer, MD, General Hospital, Belfort; Charles Santre, MD, Hospital Center, Anneey; Philippe Sauder, MD, Civilian Hospital, Strasbourg; Jean-Pierre Sollet, MD, General Hospital Victor Dupouy, Argenteuil; Thierry Soupison, MD, Hospital Center, Eaubonne; Francois Thaler, Foch Hospital, Suresnes; Franck Thomas, MD, Hospital des Diaconesses, Paris; Remi Thomas, MD, Pontchaillou Hospital, Rennes; Laurent Tronchon, MD, Hospital Center, Lens; and Michel Wolff, MD, Bichat Hospital, Paris.

Tunisia

Fekri Abroug, MD, Fattouma Bourguiba Hospital, Monastir; Moldi Amamou, MD, Emergency Medical Assistance Center, Tunis; Salah Belakhal, MD, La Rabta Hospital, Tunis; Mohammed Salah Ben Ammar, MD, M. Slim Hospital, La Marsa; Moadh Ben Miled, MD, H. Thameur Hospital, Tunis; Mohammed Besbes, MD, Ariana Hospital, Tunis; Mounir Bouaziz, MD, Habib Bourguiba Hospital, Sfax; Slah Bouchoucha, MD, Farhat Hached Hospital, Sousse; Mohamed Dhahri, MD, Military Hospital, Tunis; Zouhair Jerbi, MD, H. Thameur Hospital, Tunis; Souheil Latrous, MD, Mahdia Hospital, Tunis; and Leila Skanani, MD, La Rabta Hospital, Tunis.

South Africa

Ingrid Engelbrecht, MD, Pretoria; Reinier Pierre Krogscheepers, MD, University of the Free State, Bloemfontein; Frans Maritz, MD, Karl Bremer Hospital, Cape Town; Gregory Paul Promnitz, MD, Olivedale Clinic, Johannesburg; Johannes Stephanus Roos, MD, Vergelegen Medi-Clinic, Somerset West; Dirkie Johanna Van Rensburg, MD, Witbank; Louis Van Zyl, MD, Worcester; Isak Hendrick Vermooten, MD, Krugersdorp; and Johann Viljoen, MD, Bloemfontein.

ACKNOWLEDGMENT: We are grateful to Christine Safran, Christelle Berthou, Francoise Dellatolas, Pierre Marechal, and Ivan Brumpt for study conception and statistical analysis.

* From the Service de Reanimation Medicale et Maladies Infectieuses (Dr. Leroy), Hopital G. Chatiliez, Tourcoing, France; Departement d'Anesthesie Reanimation Chirurgicale (Dr. Saux), Hopital Sainte Marguerite, Marseille, France; Service de Reanimation Medico-Chirurgicale (Dr. Bedos), Hopital A. Miguot Le Chesnay, France; and Laboratoire Aventis (Dr. Caulin), Paris, France.

REFERENCES

(1) Niederman MS, Mandell LA, Anzueto A, et al. American Thoracic Society: guidelines for the management of adults with community-acquired pneumonia; diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med 2001; 163:1730-1754

(20 Davis R, Bryson HM. Levofloxacin: a review of its antibacterial activity, pharmacokinetics and therapeutic efficacy. Drugs 1994; 47:677-700

(3) Yamane N, Jones RN, Frei R, et al. Levofloxacin in vitro activity: results from an international comparative study with ofloxacin and ciprofloxacin. J Chemother 1994; 6:83-91

(4) Kitzis MD, Goldstein FW, Miegi M, et al. In-vitro activity of levofloxacin, a new fluoroquinolone: evaluation against Haemophilus influenzae and Moraxella catarrhalis. J Antimicrob Chemother 1999; 43(suppl):21-26

(5) Soussy CJ, Cluzel M, Ploy MC, et al. In-vitro antibacterial activity of levofloxacin against hospital isolates: a multicentre study. J Antimicrob Chemother 1999; 43(suppl):43-50

(6) Thomson KS, Chartrand SA, Sanders CC, et al. In-vitro activity of levofloxaein against Streptococcus pneumoniae with various levels of penicillin resistance. J Antimicrob Chemother 1999; 43(suppl):15-19

(7) Fish DN, Chow AT. The clinical pharmacokinetics of levofloxacin. Clin Pharmacokinet 1997; 32:101-119

(8) File TM, 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; 41:1965-1972

(9) Norrby SR, Petermann W, Willcox PA, et al. A comparative study of levofloxacin and ceftriaxone in the treatment of hospitalized patients with pneumonia. Scand J Infect Dis 1998; 30:397-404

(10) Fogarty CM, Sullivan JG, Chattman MS, et al. Once a day levofloxacin in the treatment of mild to moderate and severe community-acquired pneumonia in adults. Infect Dis Clin Pract 1998; 7:400-407

(11) Carbon C, Ariza H, Rabie WJ, et al. Comparative study of levofloxacin and amoxicillin/clavulanic acid in adults with mild-to-moderate community-acquired pneumonia. Clin Microbiol Infect 1999; 5:724-732

(12) Frank E, Liu J, Kinasewitz G, et al. A multicenter, open-label, randomized comparison of levofloxacin vs azithromycin plus ceftriaxone in hospitalized adults with moderate to severe community-acquired pneumonia. Clin Ther 2002; 8:1292-1308

(13) European Study on Community-acquired Pneumonia Committee. Guidelines for management of adult community-acquired lower respiratory tract infections. Eur Respir J 1998; 11:986-991

(14) Bartlett JG, Dowell SF, Mandell LA, et al. Practice guidelines for the management of community-acquired pneumonia in adults: Infectious Diseases Society of America. Clin Infect Dis 2000; 31:347-382

(15) Mandell LA, Marrie TJ, Grossman RF, et al. Canadian guidelines for the initial management of community-acquired pneumonia: an evidence-based update by the Canadian Infectious Diseases Society and the Canadian Thoracic Society: the Canadian Community-Acquired Pneumonia Working Group. Clin Infect Dis 2000; 31:383-421

(16) Revision de la [IV.sup.e] Conference de Consensus en Therapeutique anti-infectieuse de la Societe de Pathologic Infectieuse de Langue Francaise. Prise en charge des infections des voies respiratoires basses. Med Mal Infect 2001; 31:268-301

(17) British Thoracic Society. Guidelines for the management of community-acquired pneumonia in adults. Thorax 2001; 56(suppl):iv1-iv64

(18) Mandell LA, Bartlett JG, Dowell SF, et al. Update of practice guidelines for the management of community-acquired pneumonia in immunocompetent adults. Clin Infect Dis 2003; 37:1405-1433

(19) Bone RC, Balk RA, Cerra FB, et al. Definition for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis: The ACCP/SCCM Consensus Conference Committee; American College of Chest Physicians/Society of Critical Care Medicine. Chest 1992; 101:1644-1655

(20) Le Gall JR, Lemeshow S, Saulnier F. Simplified Acute Physiology Score: a new simplified acute physiology score (SAPS II) based on European/North American multicenter study. JAMA 1993; 270:2957-2963

(21) Bernard GR, Artigas A, Brigham KL, et al. The American-European consensus conference on ARDS. Am J Respir Crit Care Med 1994; 149:818-824

(22) US Food and Drug Administration. Coding symbols for thesaurus of adverse reactions terms. 5th ed. Rockville, MD: US Food and Drug Administration, 1995

(23) Committee for Proprietary Medicinal Products, the European Agency for the Evaluation of Medicinal Products. Note for guidance on evaluation of new anti-bacterial medicinal products, London, UK: European Agency for the Evaluation of Medicinal Products, April 1997; Publication CPMP/EWP/ 558/95

(24) Committee for Proprietary Medicinal Products, the European Agency for the Evaluation of Medicinal Products. Points to consider on switching between superiority and non-inferiority. London, UK: European Agency for the Evaluation of Medicinal Products, July 2000; Publication CPMP/EWP/ 482/99

(25) Leroy O, Santre C, Beuscart C, et al. A five-year study of severe community-acquired pneumonia with emphasis on prognosis in patients admitted to an intensive care unit. Intensive Care Med 1995; 21:24-31

(26) Torres A, Serra-Batlles J, Ferrer A, et al. Severe community-acquired pneumonia: epidemiology and prognostic factors. Am Rev Respir Dis 1991; 144:312-318

(27) Rello J, Catalan M, Diaz E, et al. Associations between empirical antimicrobial therapy at the hospital and mortality in patients with severe community-acquired pnemnonia. Intensive Care Med 2002; 28:1030-1035

(28) Ruiz M, Ewig S, Torres A, et al. Severe community-acquired pneumonia: risk factors and follow-up epidemiology. Am J Respir Crit Care Med 1999; 160:923-929

(29) Salkind AR, Cuddy PG, Foxworth JW. Fluoroquinolone treatment of community-acquired pneumonia: a meta-analysis. Ann Pharmacother 2002; 36:1938-1943

(30) Fogarty C, Siami G, Kohler R, et al. Multicenter, open-label, randomized study to compare the safety and efficacy of levofloxacin versus ceftriaxone sodium and erythromycin followed by clarithromycin and amoxicillin-clavulanate in the treatment of serious community-acquired pneumonia in adults. Clin Infect Dis 2004; 38:S16-S23

(31) 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; 148:1418-1426

(32) Ewig S, Ruiz M, Mensa J, et al. Severe community-acquired pneumonia: assessment of severity criteria. Am J Respir Crit Care Med 1998; 158:1102-1108

(33) Pachon J, Prados MD, Capote F, et al. Severe community-acquired pneumonia: etiology, prognosis and treatment. Am Rev Respir Dis 1990; 142:369-373

(34) Rello J, Bodi M, Mariscal D, et al. Microbiological testing and outcome of patients with severe community-acquired pneumonia. Chest 2003; 123:174-180

(35) Moine P, Vercken JB, Chevret S, et al. Severe community-acquired pneumonia: etiology, epidemiology, and prognosis' factors; French Study Group for Community-Acquired Pneumonia in the Intensive Care Unit. Chest 1994; 105:1487-1495

(36) Angus DC, Marrie TJ, Obrosky DS, et al. Severe community-acquired pneumonia: use of intensive care services and evaluation of American and British Thoracic Society Diagnostic criteria. Am J Respir Crit Care Med 2002; 166:717-723

(37) Leroy O, Devos P, Guery B, et al. Simplified prediction rule for prognosis of patients with severe community-acquired pneumonia in ICUs. Chest 1999; 116:157-165

(38) Kahn JB, Bahai N, Wiesinger BA, et al. Cumulative clinical trial experience with levofloxacin for patients with community-acquired pneumonia-associated pneumococcal bacteremia. Clin Infect Dis 2004; 38:S34-S42

(39) Mufson MA, Stanek RJ. Bacteremic pneumococcal pneumonia in one American city: a 20-year longitudinal study, 1978-1997. Am J Med 1999; 107(supp1): 34S-43S

(40) Waterer GW, Somes GW, Wunderink RG. Monotherapy may be suboptimal for severe bacteremic pneumococcal pneumonia. Arch Intern Med 2001; 161:1837-1842

(41) Martinez JA, Horcajada JP, Almela M, et al. Addition of a macrolide to a beta-lactam-based empirical antibiotic regimen is associated with lower in-hospital mortality for patients with bacteremic pneumococcal pneumonia. Clin Infect Dis 2003; 36:389-395

Olivier Leroy, MD; Pierre Saux, MD; Jean-Pierre Bedos, MD; and Evelyne Caulin, MD; for the Levofloxacin Study Group ([dagger])

([dagger]) A list of the members of the Levofloxacin Study Group is located in the Appendix.

This study was supported by a grant from Aventis, Paris, France. Dr. Canlin is an employee of Aventis. Drs. Leroy, Sanx, and Bedos are consultants for and have received honoraria from Aventis for lectures.

Manuscript received March 15, 2004; revision accepted December 1, 2004.

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml).

Correspondence to: Olivier Leroy, MD, Service de Reanimation Medicale et Maladies Infectieuses, Hopital G. Chatiliez, 135 rue du President Coty, 59208 Tourcoing, France; e-mail: oleroy@chtourcoing.fr

COPYRIGHT 2005 American College of Chest Physicians
COPYRIGHT 2005 Gale Group