In Reply We thank Dr Garden1 for his interest and commentary. He brings up 2 general study limitations: the surgical and nonsurgical treatment groups have certain inherent differences, and the nonsurgical treatment paradigms were heterogeneous. These are standard limitations of nonrandomized retrospective research studies. We performed multivariable analyses with propensity weighting to attempt to account for these differences. However, we agree with both of these concerns as delineated in the Discussion section of the manuscript. This is a hypothesis-generating study rather than a hypothesis-proving study.