Letters Section Editor: Stephen J. Lurie,
MD, PhD, Senior Editor.
In Reply: We agree with Drs Karnath and Luh
that our study did not include the full complement of clinical factors that
might warrant broad-spectrum antibiotic treatment. However, there is no evidence
to support their assertion that broad-spectrum antibiotics improve outcomes
in patients with ARTIs and recent antibiotic treatment or a history of smoking.
We also agree that some of the visits included in our analysis could have
represented return visits, presumably for treatment failure. However, this
proportion was likely to be small,1 and
with the exception of acute otitis media (which represents a very small proportion
of ARTIs in adults) we do not believe a return visit routinely requires treatment
with a broad-spectrum antibiotic. In fact, most cases of "treatment failure"
are likely due to use of an ineffective therapy for a viral infection (eg,
antibiotics), as it is difficult to accurately identify the small minority
of patients whose ARTIs actually have a bacterial etiology.2 Finally,
Karnath and Luh are concerned about potential confounding factors. By definition,
however, confounding factors are only relevant to the extent that they are
differentially distributed across levels of a second variable. Karnath and
Luh do not provide any rationale for suspecting that smoking behaviors, comorbidities,
or antecedent treatment failures would be more common for patients cared for
by internists compared with family physicians, or that they vary between geographic
Steinman MA, Landefeld CS, Gonzales R. Acetylcysteine and Renal Function Following Coronary Angiographic Procedures—Reply. JAMA. 2003;289(21):2796-2798. doi:10.1001/jama.289.21.2796-a