[Skip to Content]
[Skip to Content Landing]
October 11, 1995

Antibiotic Treatment for Exacerbations of Chronic Obstructive Pulmonary Disease-Reply

Author Affiliations

University of California, San Francisco

JAMA. 1995;274(14):1131-1132. doi:10.1001/jama.1995.03530140043021

In Reply.  —Dr Gilkey raises several important issues regarding our meta-analysis. First, he is concerned that by our grouping of studies that used different clinical outcomes, oranges may have been compared with apples. We agree that when all outcomes are measured on the same scale, metaanalysis is much more straightforward. However, analyses based on units of standard deviation are widely reported and, when performed properly, the results can often help clarify medical uncertainty. Given the diversity of outcome measures used in the nine randomized trials that satisfied inclusion criteria, the only meaningful way to compare antibiotics with placebo was to use the statistically acceptable effect size measurement. Each outcome was transformed into a common metric, a unit of standard deviation, before we performed our analysis. Fortunately, six of the studies included used peak expiratory flow rate as an outcome measure; thus, we were able to provide some quantitative assessment of

First Page Preview View Large
First page PDF preview
First page PDF preview