© 2005 American Thoracic Society
Early Therapy in Chronic Obstructive Pulmonary DiseaseSelection Bias by Exclusion?To the Editor:The study by Wilkinson and colleagues (1) on the benefit of early therapy of chronic obstructive pulmonary disease exacerbations poses an interesting methodologic question. The study found that the earlier an exacerbation is treated by a physician, the shorter the recovery time. This finding is best visually expressed in Figure 3 of the article by the plot of the delay between onset and treatment against the time from treatment to recovery. This figure is based on 625 treated exacerbations. There were, however, an additional 33 exacerbations that were seen by a physician but did not involve treatment, and were excluded from the analysis. Naturally, they were excluded because they were not treated and therefore could not provide data on the time from onset to treatment. The intriguing methodologic question is whether these exacerbations should and could in fact be included in the analysis. One possible interpretation of these exacerbations is that the patients delayed seeking treatment and, by the time they were ready to see their physician, their exacerbation had improved. Consequently, these patients could be included in the analysis as patients whose time from exacerbation onset to treatment would be equal to the time from exacerbation onset to recovery. Thus, the treatment recovery time for these patients could be considered to be zero. These 33 exacerbations would therefore fill the lower part of Figure 3 along the horizontal axis. It would be quite informative to recompute the regression coefficient with these 33 exacerbations included in this way. I would note that the labels and legend of Figure 3 are confusing. The legend refers to "total recovery time," yet the label uses "symptom recovery time." The latter term is a new term that pops up suddenly and is not presented in the definitional Figure 1. Moreover, it appears that this time resembles what is called "treated recovery time" and not the "total recovery time" mentioned in the legend.
McGill University Montreal, Quebec, Canada FOOTNOTES Conflict of Interest Statement: S.S. has been reimbursed for attending several conferences and also participated as a speaker in scientific meetings financed by various pharmaceutical companies: Schering-Plough, AstraZeneca, and GlaxoSmithKline. In addition, he received funding for research grants from AstraZeneca ($89,000), Schering-Plough ($80,000), and GlaxoSmithKline ($159,000). REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
