© 2008 American Thoracic Society
Inhaled Corticosteroids Might Not Protect against Lung CancerFrom the Authors:We appreciate the comments on our article (1) by Dr. Chang and colleagues. We certainly agree that influential data points may distort a true relationship. However, this effect is often less pronounced when continuous exposure variables are transformed into categorical variables. Typically, the amount of information in these transformations is reduced. We analyzed the data using both continuous and categorical classifications and found similar results: individuals receiving higher doses of inhaled corticosteroids were at reduced risk of lung cancer. In addition, we made an a priori decision to analyze our data both continuously and categorically and that our categorical variable cut points would be based on a distribution to give us approximately an equal number of exposed individuals in each stratum. In regard to the follow-up time, we restricted enrollment to individuals who did not have evidence of lung cancer; however, it remains distinctly possible that some disease was included. We would like to emphasize that even after restricting our analyses to those individuals who had a lung cancer diagnosis after 1 year of enrollment, our principal effect persisted. In addition, our analyses of confounding by indication suggested that individuals who had lung cancer were more likely, not less likely, to receive an inhaled corticosteroid. We appreciate the reframing of our results in Chang and colleagues' Table 1. However, the results are identical to those that we presented and we believe readily interpretable in any form. The history of lung cancer prevention studies suggests that all clinicians should be cautious (2, 3) not only when interpreting the results of observational studies, but also should apply equal care in the setting of clinical trials where approximately 40% of the participants not randomized to inhaled corticosteroids may have received this medication after withdrawal (4).
VA Puget Sound Health Care System
Fred Hutchinson Cancer Research Center
VA Puget Sound Health Care System FOOTNOTES Conflict of Interest Statement: D.H.A. was compensated $1,500 in 2006 from GlaxoSmithKline for participating in the "Assessing the Impact of Recent Label Updates for ADVAIR and Serevent Special Issues Board." J.W.C. has no financial relationship with a commercial entity that has an interest in the subject of this manuscript. C.L.B. has no financial relationship with a commercial entity that has an interest in the subject of this manuscript. REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
