To the Editor: We disagree with Dr. Case's view that when no difference is found between two potentially active treatments, confidence intervals would inform the question of whether a study demonstrates that the treatments are equivalent. Confidence intervals could not be used to answer the question of equivalence because our investigation was not designed or powered for an equivalence study. The lack of statistical significance, even with narrow confidence intervals, should not be used to conclude equivalence. It should be used only to conclude lack of difference. We agree that additional information, particularly confidence intervals, would have been helpful within our original report. Because of space limitations, we omitted this information. However, we have now provided the confidence intervals as a data supplement. Finally, upon examining our results, even if the study were much larger and had sufficient power to detect a difference, that difference would not have been clinically important.