Letters Section Editor: Stephen J. Lurie,
MD, PhD, Senior Editor.
To the Editor: In their study of broad-spectrum
antibiotics to treat ARTIs, Dr Steinman and colleagues1 pooled
data from 3 years of the National Ambulatory Medical Care Survey (NAMCS).
We are concerned about the validity of the estimates yielded by this data
set. The authors used the weights available in the NAMCS and weighted the
data in an attempt to make nationally representative estimates. The sampling
frame for the NAMCS and corresponding weights allow computation of estimates
for each individual year. Thus, the weights are sample-specific for each year.
Because Steinman et al pooled 3 years of unique survey data, they thereby
removed the probability-based nature of the data. Even with the sampling weights,
Steinman et al did not analyze the data in a way that would provide national
estimates of prescribing behaviors.
Mainous III AG, Hueston WJ. Variation in Use of Broad-Spectrum Antibiotics for Acute Respiratory Tract InfectionVariation in Use of Broad-Spectrum Antibiotics for Acute Respiratory Tract Infection. JAMA. 2003;289(21):2796-2798. doi:10.1001/jama.289.21.2796-a