Development Of A Multivariate Formula For Estimating The Spherical Equivalent In Refractive Surgery Candidates Using Biometric And Demographic Data
Published 2025 - 43rd Congress of the ESCRS
Reference: PO686 | Type: Free paper | DOI: 10.82333/vtmk-4b93
Authors: Arie Marcovich* 1
1Ophthalmology,Kaplan Medical Center ,Rehovot,Israel;Faculty of Medicine,Hebrew University,Jerusalem,Israel;EyeYon Medical,Nes Ziona,Israel
Purpose
This study aims to develop a multivariate formula for estimating the spherical equivalent (SEQ) in patients undergoing refractive surgery using biometric data. Additionally, it evaluates whether SimKavg or Mean Pupil Power (MPP) at 3 mm provides a more accurate SEQ estimation. The majority of the study population consisted of myopic patients.
Setting
Sekal Microchirurgia Rovigo, Rovigo Italy
Methods
A retrospective analysis was conducted on 163 eyes of young patients with clear lenses, forming the development dataset, and 50 additional eyes for validation. The collected biometric parameters included age, SimKavg, mean pupil power at 3 mm (MPP), anterior chamber depth (ACD), lens thickness (LT), axial length (AL), and cycloplegic SEQ. Axial length was converted into CMAL for standardization. A multivariate regression model was developed and validated using statistical analysis, including Pearson correlation and regression performance evaluation through the coefficient of determination (R²) and mean prediction error.
Results
The study population had a mean age of 30.24 ± 10.99 years, with SimKavg and MPP averaging 43.63 ± 1.29 D and 43.06 ± 1.36 D, respectively. ACD measured 3.59 ± 0.40 mm, lens thickness 4.01 ± 0.42 mm, and axial length 24.48 ± 1.88 mm. The real SEQ was -2.29 ± 4.72 D.
The predicted SEQ using SimKavg was -2.32 ± 4.44 D, while using MPP it was -2.35 ± 4.47 D, with respective p-values of 0.84 and 0.67. Pearson correlation coefficients showed R² = 0.9686 (p < .01) for SimKavg and R² = 0.9612 (p < .01) for MPP. However, SEQ prediction using SimKavg was within 1D of error in 84% of cases, whereas using MPP, it was within 1D in only 76% of cases (p < .01).
Conclusions
Both SimKavg and MPP appear to be valuable for predicting SEQ, assisting ophthalmologists in obtaining more accurate SEQ measurements and reducing refractive errors during consultations. However, SimKavg provided a more precise prediction, making it the more reliable parameter for clinical practice.