Community-acquired pneumonia (CAP) continues to be a topic of great interest to physicians, but one that remains filled with controversy because of the limitations of current diagnostic testing and therapeutic options. Unfortunately, Brown and Lerner's letter serves only to emphasize the emotion and misunderstanding in this field. Their letter is unnecessarily critical and inaccurate and is typical of the controversy in this area. It certainly does not accurately reflect the content of our Update or the ATS guidelines on CAP [1].

Brown and Lerner ask how Bartlett and Mundy's review could be interpreted to mean that the ATS guidelines for CAP are appropriate. The Update clearly states that Bartlett and Mundy recommended routine diagnostic testing in CAP, a position that differs from the ATS guidelines. It then stated that on the basis of the pathogens causing CAP (and shown in Table 3 of the Update), the ATS guidelines were appropriate. In fact, the pathogens in the Table were almost identical to the pathogens that were reported in the ATS guidelines to cause CAP in hospitalized patients. On the basis of these data, Bartlett and Mundy recommended empirical therapy (when necessary) for hospitalized patients that is identical to that recommended in the ATS guidelines [2]. This was the only point being made in the Update. However, in another paper not discussed in the Update, Bartlett and Mundy specifically stated that on the basis of the bacteriology of CAP seen in immunocompetent patients in their hospital, their findings "support the current American Thoracic Society guidelines for selective use of macrolide therapy" [3].

Brown and Lerner's comments seem designed to stimulate controversy rather than to emphasize points of agreement between the ATS guidelines and other approaches. Much of this stems from an apparent failure to read what the ATS guidelines say. Contrary to Brown and Lerner's statement that "the ATS guidelines advocate empirical therapy for all patients," the guidelines actually say "if a specific organism is identified, then treatment ... can be specifically directed against this pathogen." However, the "clinician often does not face such a straightforward decision and ... initial empiric therapy is by necessity somewhat broader in spectrum than is perhaps necessary. Obviously, once more information becomes available ... appropriate modifications can be made" [1].

One statement by Brown and Lerner is hard to argue with; that we can do better. It is time to stop the silly arguments about the value of specific diagnostic studies in CAP and to show whether their use leads to any measurable effect on outcome or antimicrobial resistance patterns. We are in the process of collecting validation data on the antibiotic recommendations in the ATS guidelines, and preliminary findings suggest that therapy according to the guidelines leads to improved patient outcome compared with other therapeutic approaches [4]. Unless Brown and Lerner have similar data obtained by use of their approach to diagnostic testing and therapy, their argument remains an unhelpful emotional appeal.