An open-label, randomized comparison of levofloxacin and amoxicillin/clavulanate plus clarithromycin for the treatment of hospitalized patients with community-acquired pneumonia

Ting-Yu Lin,Shu-Min Lin,Hao-Cheng Chen,Chih-Jan Wang,Yu-Min Wang,Min-Li Chang,Chun-Hua Wang,Chien-Ying Liu,Horng-Chyuan Lin,Chih-Ten Yu,Ling-Ling Hsieh,Han-Pin Kuo,Chien-Da Huang
Abstract:Background: Anti-pneumococcal fluoroquinolone has been used to treat community-acquired pneumonia (CAP) frequently because of its broad antimicrobial spectrum. Methods: This randomized, open-label study was conducted in a tertiary teaching hospital. Eligible patients were randomized to levofloxacin 500 mg IV q24h followed by 500 mg orally q24h or a combination of amoxicillin/clavulanate 500 mg/100 mg IV q8h with oral clarithromycin 500 mg q12h and then oral amoxicillin/clavulanate 250 mg/125 mg q8h with oral clarithromycin 500 mg q12h for 7-14 days. Results: From July 2004 to February 2006, 50 patients were enrolled (levofloxacin, n = 26; combination therapy, n = 24). The clinical response rate in the clinically evaluable population was similar for both groups (78.3% vs. 77.3%; p = 1.000). Levofloxacin had a higher microbiological response rate overall, and for Gram-negative and non-pseudomonas Gram-negative pathogens than the combination therapy but the difference was not statistically significant (60.0% vs. 38.9%, 55.0% vs. 21.0% and 75.0% vs. 25.0%, respectively). The length of hospital stay was similar for both groups (7.4 +/- 3.1 vs. 6.8 +/- 2.1 days; p = 1.000). Conclusion: Patients who were admitted to our hospital for CAP were older and had more comorbidities with a much higher incidence of Gram-negative pathogens than in a previous study. Levofloxacin was at least as effective as amoxicillin/clavulanate plus clarithromycin in clinical and microbiological responses. Levofloxacin had a higher microbiological eradication rate than the combination therapy but the difference was not statistically significant. This deserves further study with a larer sample size.
What problem does this paper attempt to address?