Official ESCRS | European Society of Cataract & Refractive Surgeons
Lisbon 2017 Delegate Registration Programme Exhibition Virtual Exhibition Satellites OneWorld Travel Discount
escrs app advert

Posterior surface corrected keratometry: comparison of Barrett calculator with other methods

Search Title by author or title

Session Details

Session Title: Pseudophakic IOLs: Toric

Session Date/Time: Tuesday 10/10/2017 | 08:30-10:30

Paper Time: 08:54

Venue: Room 4.4

First Author: : P.Hoffmann GERMANY

Co Author(s): :    M. Abraham                    

Abstract Details


Recently, the impact of the posterior surface on total corneal astigmatism has come into focus among cataract and refractive surgeons. While it would be best to have precise measurements of both corneal surfaces and appropriate software to deal with the data, most surgeons still rely on time proven keratometry. Different algorithms exist for estimating the impact of the posterior curvature on total corneal astigmatism. One of the best known is the Barrett toric calculator. We evaluated this software in comparison to other methods.


Augen- & Laserklinik Castrop-Rauxel


A cohort of 104 low to moderately astigmatic eyes that had been previously implanted with rotationally symmetric IOLs was selected. All eyes were measured with Lenstar automated keratometry as well as Casia SS-1000 swept source anterior segment OCT. Lenstar data was used as is, Barrett corrected and also “Castrop corrected” with an algorithm developed in-house in 2014. A subjective refraction was taken and the difference vector (DV) between subjective and objective cylinder was calculated. DV was a measure of how precise cylindrical refraction can be predicted from preoperative corneal data.


For regular keratometry, DV was 0.56 ± 0.27 D. When Lenstar data was corrected with the Barrett toric calculator, DV was 0.41 ± 0.28 D. The Castrop algorithm yielded 0.39 ± 0.25 D. A real measurement with the OCT delivered a DV of 0.44 ± 0.22 D. The number of outliers > 1 D was 5.8% for Lenstar, 4.8% for Barrett, 2.9% for Castrop and 0% for OCT. The differences between Lenstar uncorrected and all other methods were statistically significant (repeated measures ANOVA) while differences between Barrett, Castrop and OCT were not.


The impact of the posterior corneal curvature on total corneal astigmatism is most pronounced in small astigmatism. The effect can be measured e.g. with a swept source OCT or Scheimpflug tomography or estimated from keratometric data. It could be shown that the popular Barrett algorithm improves the predicitive precision of high-quality keratometry significantly. The same is true for the very simple Castrop algorithm. Both are as effective as an expensive OCT measurement. While we would advocate to do both before implanting a toric IOL, a single keratometry and appropriate algorithmic correction would give a satisfying result in >95% of eyes.

Financial Disclosure:


Back to previous