Dr Carones et al1 concluded in their article published in the July 1994 issue of the Archives that photorefractive keratectomy caused no damage to the corneal endothelium. This conclusion was reached after comparing preoperative with postoperative values of cell density, coefficients of cell area variation, and percentages of hexagonal cells and after testing for a linear association of endothelial cell loss, coefficient of cell area variation, and percentage of hexagonal cells with depth of photoablation (ie, Pearson's correlation coefficient). However, we would like to point out a methodological problem in their article that could invalidate their conclusions.
The fact that they did not find any linear association (ie, correlation coefficients of.23 or less) does not necessarily mean there is no association.2 Dr Carones et al should have noted that a correlation coefficient could be zero and there still could be a nonlinear relation between the two variables.
Soto-Pedre E, Hernéz-Ortega C. The Corneal Endothelium After Myopic Excimer Laser Photorefractive Keratectomy. Arch Ophthalmol. 1995;113(11):1356. doi:10.1001/archopht.1995.01100110016007