We would like to thank Richoz and associates for their interest in our trial. We disagree with their observation that the main conclusion of our trial was an increased risk of perforations in the cross-linking group. We explictly stated that the prespecified primary outcome of the trial was treatment failure, defined as perforation and/or an increase in infiltrate size by ≥2 mm. The main conclusion of the trial was the lack of a significant difference in treatment failure between the cross-linking and control groups, which was clearly stated as the primary result in the abstract and paper. Despite the small sample size, we nonetheless found a significant increase in the secondary outcome of perforation in the cross-linking arm. The authors suggest that this could have happened by chance, and we agree. If we take into account all 5 P values reported in the results section and assume that each was an independent outcome with a significance level of .05, then this leaves a 25% chance of finding one of these P values to be significant by chance alone. Although the small sample size is a limitation, we thought the readers would still be interested in the results of this small randomized controlled trial as they decide whether to pursue cross-linking for their patients with recalcitrant deep fungal ulcers. We specifically noted the small sample size as a limitation of the trial in several places in the manuscript, including the concluding sentences of the paper.

The authors seem to suggest that the small sample size led to imbalanced randomization, which is a possibility with small samples. As stated in our paper, we did not find any significant differences in the baseline characteristics between the 2 groups. We agree with the authors that the reader should keep in mind the small sample size and potential for imbalance, though we do not believe we found evidence to suggest that the groups were actually imbalanced.

Another limitation raised by the authors was our use of the Dresden protocol, which they state may not be effective for deep ulcers. We used the Dresden protocol because it had been used in previous studies of corneal ulcers, but we agree that alternative protocols should be considered for cross-linking deeper stromal ulcers. We disagree that the study had fundamental experimental flaws or that we made inadequately evidenced assertions; certainly the trial had limitations, but we tried to be cautious in our conclusions and to remind the reader to interpret the results in light of the small sample size. As we stated in our paper, we do not believe that this single trial means that cross-linking has no role in the treatment of infectious keratitis, and we encourage further study on the topic.