Author’s response: influenza-associated mortality oseltamivir: beware of misstepping into stepwise procedures

摘要

We thank Lytras et al. [1] for the comments on our recent article assessing risk factors for mortality in inpatients with influenza and the effect of oseltamivir [2]. For this study, the authors are confident that the identification and adjustment for confounding variables have been done in a systematic and objective manner using step- wise logistic regression. We note the concerns on the methods we have used, but we do not see how these concerns are specific to our work. Rather the concerns relate in general to the application of these methods.