A Simplified Method To Minimize Systematic Bias Of Single-Optimized Intraocular Lens Power Calculation Formulas.
Published 2023 - 41st Congress of the ESCRS
Reference: PP22.03 | Type: Free paper | DOI: 10.82333/a7pn-yv95
Authors: Damien Gatinel* 1 , Guillaume Debellemanière 1 , Alain Saad 1 , Radhika Rampat 1 , Avi Wallerstein 2 , Mathieu Gauvain 2 , Jacques Malet 1
1Rothschild Foundation Hospital,Paris,France, 2Ophthalmology and Visual Sciences,McGill University ,Montréal ,Canada
Purpose
To provide a simplified method to optimize lens constants to zero the Mean Prediction Error (ME) of an intraocular (IOL) calculation formula, without the need to program the formula itself, by exploring the influence of IOL and corneal power on the refractive impact of variations in effective lens position.
Setting
Theoretical development and retrospective clinical evaluation on documented datasets.
Retrospective data from 8878 patients with cataracts with pre- and postoperative measurements available using four IOL models and six IOL power calculation formulas.
Methods
A schematic eye model was used to study the impact of small variations in effective lens position (ELP) on the postoperative spherical equivalent (SE) refraction. The impact of keratometry (K) and IOL power (P) on SE was investigated. A theoretical thick lens model was used to devise a formula to zero the average prediction error of an IOL power calculation formula. This was achieved by incrementing the predicted ELP, which could then be translated into an increment in the IOL constant. This method was tested on documented real-life postoperative datasets, using different IOL models and single-constant optimized IOL calculation formulas.
Results
: For small variations in ELP, there is an exponential relationship between IOL power and the resultant postoperative refractive variation. The ELP adjustment necessary to zero the MPE equals a ratio between the ME and the mean of the following expression: 0.0006*(P²+2KP) on the considered datasets. The accuracy of the values obtained using this formula was confirmed on documented postoperative datasets, and on published and non-published formulas.
Conclusions
The proposed method allows surgeons without special expertise to optimize an IOL constant to nullify the ME on a documented dataset without coding the different formulas. The influence of individual eyes is proportional to the squared power of the implanted IOL.