A Comparative Analysis Of Accuracy Of 7 Intraocular Lens Power Calculation Formulas Available On Escrs Iol Calculator
Published 2024 - 42nd Congress of the ESCRS
Reference: FP06.01 | Type: Free paper | DOI: 10.82333/hqv4-ae79
Authors: Twinkle Shah* 1 , Rupal Shah 2 , Abhishek Shah 2 , Kaushambi Ranpara 3
1Ophthalmology,Manish Eye Institute,Ahmedabad,India, 2Ophthalmology,M&J Institute,Ahmedabad,India, 3Ophthalmology,Geetanjali medical college and hospital,Udaipur,India
Purpose
To evaluate the accuracy of 7 IOL power formulas available on open source ESCRS IOL calculator (Barrett, Cooke K6, EVO, Hill-RBF, Hoffer QST, Kane and Pearl DGS) in predicting the post operative refractive error.
Setting
A retrospective case study done at a private practice clinic in Ahmedabad, Gujarat, India.
Methods
Data was collected for 95 eyes of 95 patients who underwent uneventful cataract surgery with Acrysof IQ SN60WF implantation by a single surgeon over 3 months. The predicted refractive outcome was calculated for 7 different formulas using the open source ESCRS calculator, using optimized lens constants. These were then compared with the actual refractive outcome to give prediction errors for each formula. Mean absolute error was the primary outcome. Eyes were divided into 3 groups: Short(≤22mm), Medium(>22to<26mm) and Long(≥26mm).
Results
Over the entire axial length range, the Kane formula had the lowest mean absolute prediction error (0.247) followed by EVO (0.254) and Barrett (0.261). Kane formula showcased the highest percentage of eyes with prediction errors between ±0.50D (i.e. 60%) followed by Pearl DGS(58.95%), Barrett(54.74%) and EVO(54.74%). Apart from Kane and EVO, Cooke K6 was a better predictor in short axial lengths with lower mean absolute prediction error of 0.318 but in normal and longer eyeballs Barrett yielded lower MAE of 0.259 and 0.198 respectively.
Conclusions
Kane formula was a more accurate predictor of actual post operative refraction than the other formulae in this study. EVO showed a better accuracy in shorter eye balls whereas Hill-RBF showed a better reliability in medium or normal range of axial lengths. Barrett worked better with medium and long axial lengths but lacked reliability in shorter axial lengths.