First, I would like to extend my gratitude to de Bernardo et al for their comments on our study. In this study, early progression of central corneal thickness (CCT) after cataract surgery was evaluated based on surgeons' experience.1 We compared patients operated on by an experienced surgeon, who had performed over 18,000 procedures, with patients operated on by less experienced surgeons who had each performed a total of 500 procedures. After adjusting for confounding factors, it seemed that the only factor significantly associated with postoperative corneal edema was surgical experience (P = 0.0005). Extensive surgical experience was associated with reduced postoperative corneal edema, shorter surgery time, and shorter phacoemulsification time.
As highlighted by de Bernardo and associates, we found noncontact pachymetry to be the most appropriate method for providing accurate comparative measurements, eliminating the risk of interobserver bias. Measurements obtained by ultrasound pachymetry are indeed more accurate but are also more operator dependent.2,3 Moreover, it requires administration of anesthetic eye drops, which in turn may cause measurements to fluctuate.4
Regarding the differences between the senior and the junior groups, the only impact of such a difference in terms of statistical calculation is that statistical power is reduced in which subject numbers are identical. However, regarding the included subject numbers, the statistical power required to demonstrate a significant difference of half the SD with an alpha risk of 0.05 was 80% and that of 1 SD was over 95%. There is nevertheless a possible selection bias regarding the distribution of patients between the senior surgeon and the junior surgeons. However, this bias exists independently of the disparity in size between both groups, and we have tried to take this into account by incorporating potential confounding factors into the multivariate analysis. Because this is a nonrandomized study, it must be acknowledged that there remains a risk of bias. Consequently, a prospective randomized study is needed to improve the degree of certainty of our conclusions, as expressed by de Bernardo and associates.
The surgical technique was not retained in the final multivariate analysis model because it was not one of the variables significantly associated with an increase in postoperative CCT using bivariate analysis. A multivariate model with the surgical technique as one of the variables proves that it is not significantly associated with an increase in corneal thickness in our study (Table 1). In a recent comparative study including 110 eyes, it was confirmed that there is no significant difference in early postoperative CCT when comparing tilt-and-tumble and divide-and-conquer techniques in cataract surgery.5 Finally, it should be noted that a prospective randomized study on a larger scale (the PERCEPOLIS Study: Endothelial cell loss after intra or supracapsular phacoemulsification) corroborating this result has recently been completed and will soon be published.