Spherical Equivalent Prediction Analysis In Iol Power Calculations Following Myopic Laser Vision Correction
Published 2025 - 43rd Congress of the ESCRS
Reference: PO454 | Type: Free paper | DOI: 10.82333/8bzw-h692
Authors: Rustu Emre Akcan* 1 , Ayşe Yıldız 1 , Cem Kesim 1 , Arda Gulersoy 2 , Koray Kavakli 2 , Ehsan Varasteh 2 , Parviz Zolfaghari 2 , Murat Hasanreisoğlu 3 , Christina Nikolova Grupcheva 4 , Hakan Urey 5 , Afsun Sahin 6
1Ophthalmology,Koç University Hospital,Istanbul,Türkiye, 2Electrical and Electronics Engineering,Koç University,Istanbul,Türkiye, 3Ophthalmology,Koç University Hospital,Istanbul,Türkiye;Research Center for Translational Medicine (KUTTAM), Koç University,Istanbul,Türkiye, 4Ophthalmology and Visual Science,Varna Medical University,Varna,Bulgaria, 5Electrical and Electronics Engineering,Koç University,Istanbul,Türkiye;Research Center for Translational Medicine (KUTTAM), Koç University,Istanbul,Türkiye, 6Ophthalmology,Koc University Hospital,Istanbul,Türkiye;Research Center for Translational Medicine (KUTTAM), Koç University,Istanbul,Türkiye
Purpose
To evaluate the prediction accuracy of intraocular lens (IOL) power calculation formulas for cataract surgery after myopic laser vision correction (LVC) using both standard keratometry (K) and posterior keratometry (PK) measurements.
Setting
Single-center academic university hospital.
Methods
Single-center retrospective case series consisting of two groups of patients who had undergone successful cataract surgery following myopic LVC. Group 1 included all eyes with standard K readings. Group 2 included only eyes where both K and PK were available. All K and PK measurements were obtained using a single swept-source optical biometer. The online analytical tool Eyetemis was used for spherical equivalent prediction error (SEQ-PE) analysis. IOL constants were optimized to compare Pearl DGS, EVO 2.0, Hoffer QST, Barrett True K (BTK) and ASCRS-mean post-refractive formulas.
Results
Group 1: Forty-seven eyes from 40 patients were included. The trimmed means of BTK and ASCRS-mean formulas significantly differed from zero (p=0.040 and p=0.017, respectively) suggesting a systematic bias in their prediction. EVO 2.0 had the best precision and accuracy (0.433 and 0.423D, respectively).
Group 2: Twenty-eight eyes from 23 patients were included. The trimmed mean of HofferQST with and without PK significantly differed from zero (p=0.020 and 0.003, respectively). EVO 2.0 with PK had the best precision (0.409D). HofferQST formula, with or without PK, showed the worst precision and accuracy (0.535 and 0.603D, respectively).
Conclusions
The EVO 2.0 formula showed the best precision and accuracy, particularly when PK data was included. However, PK data did not significantly enhance the precision or accuracy of formulas.