We would like to thank Carrie Huisingh and Gerald McGwin for their interest in our paper. We have to state that we strongly disagree with most of their comments.
First, they suggest that “cohort study” defines our study better than “case series.” This is not correct. A “cohort study” is defined as “the study of a population exposed to a risk factor which may influence the occurrence of a given disease or another outcome.” This is clearly not the case of our study, which analyzes the effect of LASEK surgery on the keratocyte population. There is not a disease or a given outcome to reach, so we believe that “case series” is, by far, a more appropriate description.
Second, they state in their letter that the description of the study design has implications in the statistical analysis. This is the first time we have heard this kind of assumption. Different studies sharing the same study design may need different statistical approaches, depending on the type of variables analyzed and the distribution of the data obtained. This fact is so evident that it needs no further comments.
Third, Huisingh and McGwin suggest that paired statistics are more appropriate for the analysis of our data. They seem to have overlooked 2 important facts of our study. The main objective of our paper is the comparison of the keratocyte population in post-LASEK eyes and normal subjects. These are clearly 2 different populations, so the use of a paired test would have been absolutely incorrect. The second objective of our study was to compare the keratocyte density in the same eyes at different time points after surgery. In this case we used a repeated-measures ANOVA test, which we believe is the right test to analyze the same parameter in the same eyes at different time points, as its own name suggests.
Fourth, we apologize for the mistake we made about the Bonferroni test. We correctly indicate in our paper that the Bonferroni correction (or adjustment) was used when appropriate, for multiple comparisons, and this is correct. Unfortunately, in the previous sentence, we should have stated that the Kolmogorov-Smirnov test, not the Bonferroni test, was used to check that the data analyzed had a normal distribution. We apologize for this mistake.
Fifth, we are very surprised by their statement that “given the nature of the data, it is unlikely that the data are normally distributed.” In several studies published in prestigious journals analyzing the corneal keratocyte population either after refractive surgery or in normal eyes, the authors have found that the data obtained were normally distributed, and consequently, parametric tests were used for the statistical analysis. In conclusion, we believe that the use of parametric tests in our study was appropriate.