Calculating for Unusual Eyes

Roibeárd O’hÉineacháin reports

Eyes with unusual optical dimensions are at a higher risk for refractive surprise following cataract surgery, but newer intraocular lens calculation formulas can reduce that risk, said Giacomo Savini MD.

“The latest generation of IOL calculation formulas reduces the risk of refractive surprises, especially if you have optimised constants,” he said. “And our wish is that new biometers will include the formulas so we can have them on the printout, which would make everything easier.”

Influence of risk factors on formulas

Dr Savini noted keratometry values of less than 42.0 D or more than 46.0 D increase the risk of refractive surprises with many of the older IOL calculation formulas. Using the SRK/T, prediction accuracy sharply diminishes as the K value deviates from average values (42.0–44.0 D), with myopic errors in eyes with steep corneas and hyperopic errors in flat corneas.

The accuracy of other formulas is also influenced by keratometry values, although to a lesser degree. In eyes with flat K, using Holladay 1 and 2 can result in a hyperopic error, but using Haigis can result in a myopic error. In eyes with steep keratometry, using Haigis and Olsen can result in a hyperopic error.

All eyes with values less than 2.75 mm or more than 3.50 mm in anterior chamber depth (ACD) are at increased risk of refractive surprises. Using the old formulas in such eyes— which don’t account for ACD—can result in myopic error in eyes with shallow ACD and hyperopic error in eyes with deep ACD. The Barrett and Haigis formulas reduce this risk because they account for ACD, as do EVO, Kane, and Hoffer QST, he explained.

In eyes with an axial length shorter than 22.0 mm, only around half of cases will achieve postoperative refraction within 0.50 D of the target value with all the older formulas. That compares to around three-fourths of eyes with more average proportions. There also appear to be no significant differences between the formulas in predictive accuracy.ⁱ

Dr Savini cited recent research showing that the latest generation formulas have greater predictive accuracy in short eyes—including Kane, Naeser 2, Olsen, VRF-G, Castrop, Okulix, and Pearl-DGS.^ii–iii

Old formulas with no axial length adjustment tend to result in hyperopic errors in long eyes and tend to result in myopic errors with axial length adjustment. Barrett, Olsen, and Haigis achieve the highest accuracy in long eyes.^iv Better outcomes with newer formulas include Barrett Universal II, EVO 2.0, and Hoffer QST.^v

“The ESCRS IOL calculator is a good option because it only requires users to enter the data once. And it will perform the calculation from seven formulas that have all shown greater accuracy than standard IOL formulas available on biometers, as well as the Barrett II formula,” Dr Savini said. “However, even with the newer formulas, and even among average eyes, one in four or five will have a postoperative error of more than half a dioptre.”

This paper was presented at the 40th Congress of the ESCRS in Milan.

Giacomo Savini MD is an ophthalmologist in private practice in Bologna, Italy, and a researcher at the GB Bietti Foundation – IRCCS in Rome, Italy. giacomo.savini@startmail.com

ⁱ Gokce et al, JCRS, 2017; 43: 892.

ⁱⁱ Voytsekhivskyy et al, Curr Eye Res, 2021; 12: 1832.

ⁱⁱⁱ Wendelstein et al, BJO, 2022; 106: 795.

^iv Melles et al, Ophthalmol, 2018; 125: 169.

^v Shammas et al, JCRS, 2022; 48(10): 1113–1120.