© 2007 American Thoracic Society
Atypical Coverage for Community-acquired PneumoniaTo the Editor:The article by Dr. Arnold and colleagues on antibiotic coverage for community-acquired pneumonia (CAP) concludes that the need for atypical coverage is strongly supported by their results (1). Our confidence in this recommendation is severely limited for the following reasons. First, two systematic reviews and meta-analyses of randomized controlled trials that compared regimens with atypical coverage to regimens without such coverage showed no difference in mortality and clinical response (2, 3). The design of the study by Arnold and coworkers is observational and retrospective. We should put more confidence in results of randomized controlled trials than in those of observational, retrospective studies. Second, patients treated for atypical pathogens are probably a priori different from patients treated with a β-lactam drug alone (4). Physicians are likely to reflect in their choice of treatment common wisdom as to the presentation of "atypical" pathogens—that is, younger patients, lower fever and leukocyte count, nonproductive cough, and certain patterns of infiltrate on the chest radiography. The potential for bias, as the authors noted, is high. Classical multivariable techniques may not be suitable to adjust for the differences between the two groups of patients. Matching by propensity (for treatment regimen) might account for some of these a priori differences. We urge the authors to perform a propensity-matched analysis and report its results. Third, the authors should conduct and report a formal analysis of interaction between treatment with atypical coverage and the presence of an atypical isolate. To strengthen their conclusion, we would expect a better outcome in patients with atypical pathogens treated with atypical coverage (again adjusted for the propensity score). In conclusion, a synthesis of randomized controlled trials showed no benefit for empirical treatments of CAP covering atypical pathogens. It is doubtful whether observational studies can address this question, given the a priori differences between the patients' groups. A randomized trial comparing a β-lactam drug to the combination of a β-lactam and a macrolide drug (the most common combination in clinical practice) has not been done and is sorely needed.
Rabin Medical Center FOOTNOTES Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
