Author reply

Abstract

We thank Rosenberg et al for their interest in our article and apologize if it was not clear in the use of an intention-to-treat analysis.The authors agree that the standard analysis should include an intention-to-treat analysis studying the outcomes according to the original random assignment.Although we may not have explicitly stated that an intention-to-treat analysis was done, in fact it was done for the treatment failures and serious complications (Table 7; total complications, 12/229 patients [5.2%]). As stated in the article, no significant difference was found between the 3 groups in occurrence of these serious complications. We thank Rosenberg et al for their interest in our article and apologize if it was not clear in the use of an intention-to-treat analysis. The authors agree that the standard analysis should include an intention-to-treat analysis studying the outcomes according to the original random assignment. Although we may not have explicitly stated that an intention-to-treat analysis was done, in fact it was done for the treatment failures and serious complications (Table 7; total complications, 12/229 patients [5.2%]). As stated in the article, no significant difference was found between the 3 groups in occurrence of these serious complications. Moxifloxacin for Bacterial KeratitisOphthalmologyVol. 115Issue 6PreviewIn their article comparing fortified tobramycin/cephazolin, ofloxacin, and moxifloxacin in the treatment of bacterial keratitis, Constantinou et al1 conclude that “all 3 antibiotic treatments were equally effective against a wide range of ocular isolates in the treatment of severe bacterial keratitis.” This conclusion can be questioned. The authors do not use an intention-to-treat analysis, because not all subjects randomized are included in the analysis of efficacy. In a randomized clinical trial comparing efficacies of ≥2 medical treatments, intention-to-treat analysis should be used. Full-Text PDF