Three of the authors respond:
We thank Andrew Morris for his interest in our meta-analysis.1 We chose mortality as our primary outcome because of the considerable mortality rate attributed to community-acquired pneumonia in several previous studies. However, the mortality rate for patients with pneumonia in several of the randomized controlled trials included in our meta-analysis was extremely low, and the mortality rate for patients in the subgroups with severe or bacteremic pneumonia was not reported. Therefore, we believe that the available data were too limited to reach a strong conclusion on this important outcome.
On the other hand, secondary outcomes such as treatment success may provide significant evidence regarding the effectiveness of antibiotic regimens. We stated in our article that macrolides and β-lactams (the comparators of fluoroquinolones in the randomized controlled trials included in the meta-analysis) are also effective for the treatment of community-acquired pneumonia. However, we believe that more data on the comparative effectiveness of various antibiotics are needed for patients with severe community-acquired pneumonia. Unfortunately, the available relevant blinded and high-quality randomized controlled trials enrolled only patients with mild or moderately severe pneumonia, a group in which the anticipated success rate would be higher for all antibiotics (compared with that in patients with severe community-acquired pneumonia). Thus, we elected to include all randomized controlled trials in our meta-analysis, an approach that is not unique in the literature; for example, the Cochrane Collaboration suggests including all the available evidence in the primary analysis and proceeding to sensitivity or subgroup analyses to refine the outcomes.2 Finally, we cannot overemphasize that physicians should also be aware of the problems related to the use of the various antimicrobial agents, including fluoroquinol ones: they may be associated with ad verse events such as arrhythmias,3 Clostridium difficile–associated diarrhea4,5 and the emergence of antimicrobial resistance.6,7
Footnotes
-
Competing interests: None declared.