Optimizing Intraocular Lens Power Calculations in Eyes with Short Axial Length

This review evaluates the challenges of intraocular lens (IOL) power calculation in eyes with short axial length (AL) and examines the impact of modern optical biometry and new generation of artificial intelligence (AI) based formulas on refractive outcomes. Short AL eyes remain challenging for accurate IOL power calculation and are consistently associated with increased prediction error and refractive surprises. This is largely due to the disproportionate effect of minor biometric errors and difficulties in estimating effective lens position in small eyes. The introduction of advanced optical biometers, such as IOLMaster 700, has significantly improved measurement precision by enhancing the accuracy of AL and anterior chamber depth assessments. Furthermore, AI-based IOL formulas, such as Kane and Prediction Enhanced by Artificial Intelligence and output Linearisation – Debellemaniére, Gatinel, Saad (Pearl-DGS), which utilize large datasets and non-linear machine learning models, have demonstrated superior refractive predictability compared to traditional formulas such as Hoffer Q and Sanders-Retzlaff-Kraff/Theoretical (SRK/T). Cooke K6 formula (K6) which incorporated prediction modification, has shown the highest prediction accuracy. However, heterogeneity in study design, biometric devices, and definition of short AL limits direct comparison between studies, and no universally accepted gold-standard formula has yet been established. In conclusion, the combination of high-precision optical biometry and AI-driven IOL formulas significantly improves refractive accuracy in eyes with short AL. Further large, multicentre prospective studies are required to establish an evidence-based gold-standard formula and to identify optimal formula for each biometric phenotype, including short AL.