We appreciate Dr. Singh's and Ms. Wagener's comments but must address several points of disagreement. First, we do not agree that our conclusions may be misleading. As discussed in our paper, trial design limitations, duration of follow-up, and prophylaxis were thoroughly evaluated by sensitivity analyses, and the conclusions of our meta-analysis remained unchanged.

Second, Dr. Singh and Ms. Wagener do not feel that the studies by Hibberd (1) and Conti (2) and colleagues can truly be considered preemptive therapy. Of interest, both of these studies have the words "preemptive ganciclovir therapy" in their titles. After considering the objectives and quality of these 2 trials, we decided that they should be included in our analysis. Nonetheless, because the patients in these studies were more intensely immunosuppressed and because of the potential differences in the design of the preemptive therapy (which Dr. Singh and Ms. Wagener pointed out), we also performed a sensitivity analysis, which was not included in our manuscript. The overall results for CMV organ disease in the preemptive analysis remained similar and statistically significant (odds ratio, 0.24 [95% CI, 0.07 to 0.85]; P = 0.014 [Mantel–Haenszel test]; P = 0.188 [Breslow–Day test]).

Third, as in any well-designed, large clinical trial or meta-analysis, the sample size for the subgroup analyses is always a potential concern; no single trial or meta-analysis is likely to be adequately powered for all subgroup analyses. The studies are powered for the group, not for the subgroups. Therefore, we agree with Dr. Singh's and Ms. Wagener's concerns about the subgroup power.

Fourth, we never stated that "preemptive therapy is not as optimal as universal prophylaxis for specified subgroups," and we did not draw conclusions on the basis of subgroup results only. What we actually said in our Discussion section is found in the final paragraph: "Our results suggest that universal prophylaxis may be the preferred method of treatment because of its reduction in post-transplantation CMV organ disease and its reduction in bacterial and fungal opportunistic infections, allograft rejection, and death. However, a prospective trial...needs to be done to confirm our findings." This statement is based on the results from the primary and secondary outcome analyses.

Fifth, Dr. Singh and Ms. Wagener state, "if trials based on the use of preemptive therapy at viral detection are considered, then the analysis of the effect of this approach on mortality in Kalil and colleagues' meta-analysis...would have included only 2 studies involving a total of only 72 patients." This is incorrect. The 2 studies in question (3, 4) included 140 patients.

In conclusion, while we agree that both universal prophylaxis and preemptive approaches are very effective in preventing CMV organ disease, the debate about the differences in the magnitude of the effect and the costs can only be brought to a close by a randomized trial.