Inter-Formula Variability in IOL Power Planning in Short and Long Eyes: A Retrospective Paired Comparison of Seven ESCRS-Hosted Formulas and the Ladas Super Formula 2.0

Le résumé fourni par la source

Purpose: To quantify the frequency, magnitude, and direction of inter-formula planning variability in intraocular lens (IOL) power calculation at the extremes of axial length, comparing seven European Society of Cataract and Refractive Surgeons (ESCRS)-hosted formulas and the Ladas Super Formula 2.0, and to identify preoperative predictors of clinically relevant disagreement. Methods: In this retrospective, single-center, paired within-eye planning study, the recommended IOL power and predicted spherical equivalent were generated for short (axial length <22 mm) and long (>26 mm) eyes from identical biometry (IOLMaster 700) across the seven formulas on three IOL platforms. A clinically relevant discrepancy (ClinRel) was defined as a within-eye range ≥1 step (≥0.5 D) in recommended power or ≥0.25 D in predicted spherical equivalent across the seven formulas. Firth-penalized logistic regression identified predictors in the long stratum. Results: A total of 1020 eyes (551 short, 469 long) of 665 patients were analyzed. On the primary platform, ClinRel occurred in 98.2% of short and 62.7% of long eyes, reproduced on all three platforms and in one-eye-per-patient analyses. Per-formula mean signed differences vs Barrett Universal II were small (-0.11 to +0.16 D), yet the seven formulas diverged by ≥2 steps in 33.8% of short eyes. Hill-RBF 3.0 was out-of-bounds in ~30% of eyes; the Ladas Super Formula 2.0 (n = 311) was not a directional outlier. In long eyes, axial length was the strongest predictor (Firth odds ratio 1.55 per 1 SD; 95% CI 1.23-1.95), with no clean threshold (Youden axial length/corneal radius ratio 3.52; sensitivity 35.5%, specificity 96.0%). Conclusions: Modern IOL power formulas frequently disagree on the recommended power in short and long eyes despite comparable published accuracy. Because postoperative outcomes were not assessed, these findings quantify planning-level variability only; disagreement should alert surgeons to planning uncertainty in these eyes.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.