The use of instruments such as the Pentacam (Oculus GmbH) to measure both corneal surfaces, as done by Borasio et al.,1 is a logical approach to power calculation in eyes that have had corneal surgery, and I believe it is an improvement over current methods. However, when Borasio et al. applied the thick lens formula to the Pentacam data with a corneal refractive index of 1.376 and found the corneal power to be about 1.3 diopters (D) lower than that found by videokeratoscopy, they chose to trust the latter. They then used statistical regression to find a new corneal refractive index that with the Pentacam data produces corneal powers that agree with videokeratoscopy. The Topcon KR-8100PA used in their study, similar to most keratometers and corneal topographers, assumes a keratometric index of 1.3375 to convert measured anterior surface curvature to corneal power. Assuming Gullstrand's corneal model, in which the refractive indices of air, cornea, and aqueous are 1, 1.376, and 1.336, respectively; where thickness = 0.5 mm; and where, in particular, the ratio of posterior to anterior radii is 6.8/7.7 = 0.883, one can show that the index 1.3375 gives the power at the posterior vertex of the cornea.2 An index of 1.3315 gives the power at the principal plane,3 which is about 0.8 D less than at the posterior vertex. Recently, Dubbelman et al.4 found the posterior/anterior ratio to be on average 0.813; ie, considerably lower than 0.883. With the new ratio, corneal power is reduced by about another 0.5 D. The true corneal power in the principal plane of real virgin corneas is thus 1.3 D less than that obtained with the index 1.3375, precisely what Borasio et al. found. Briefly, this means the Pentacam results are correct and the keratometers and topographers are wrong. However, it is not possible to use the correct corneal powers for power calculations because the current formulas were based on erroneous powers that were obtained with instruments that applied the index 1.3375. The data on which the formulas were developed could be reanalyzed to correct them, but then all existing intraocular lens (IOL) constants would have to be reassigned. Exact ray tracing using Pentacam topography, proper refractive indices, precise axial length (including corneal and retinal thickness), and exact IOL design agrees well with refractive outcome. I did the power calculation for my own right eye using this method, and the agreement with the postoperative outcome was perfect. (Of course it will take more than 1 eye to prove my point, and I am working on that.) There is nothing mysterious about optical theory and ray-tracing calculations. The science is well established and applies to all optical systems, including eyes. It may be hard to understand and carry out, but programs as well as the computing power needed for the calculation are available. It is even more important to find better algorithms for predicting the postoperative position of the IOL. (For my own eye, I used a new algorithm, which I will publish in the proper context.) Has the time come to call for a moratorium on IOL power calculation using paraxial (thin or thick lens) calculation and IOL constants? I personally believe so. In my opinion, a moratorium is the only way to resolve the problems of power calculation in eyes that have had corneal refractive surgery. These eyes have considerably more aberrations, in particular, spherical aberration, that influence refractive outcome, but paraxial calculation cannot take aberrations into account. The effect of spherical aberration on power calculation can be illustrated by the AMO CeeOn 911A and Tecnis Z9000 IOLs. They have exactly the same design except for the modified prolate front surface of the Tecnis IOL and hence occupy the same position in the eye. Yet, their labeled A-constants are 118.3 D and 119.0 D, respectively, and their ACD-constants are 5.14 mm and 5.55 mm, respectively. The power calculation formulas must thus be manipulated as if the Tecnis lens were 0.41 mm deeper in the eye, which results in choosing, on average, a 0.7 D stronger Tecnis IOL to obtain the desired refractive outcome. I encourage Borasio et al. and all others who struggle with post-keratorefractive IOL power calculation to work further in the direction of making exact measurements and applying exact calculations that are based on established physical principles. In the meantime, we are limited to such semi-empirical approaches as the BESSt formula, which indeed appears to be an improvement over the current methods and for which its authors should be commended.
No takes yet. Share an insight, caveat, or question.
Sverker Norrby (2008) studied this question.