Comparison of IOL Power Calculation Formulas After Sutureless Intrascleral One-Piece IOL Fixation: A Post-hoc Analysis.

PURPOSE: To compare the refractive prediction accuracy of five intraocular lens (IOL) power calculation formulas in eyes undergoing secondary sutureless intrascleral fixation of a one-piece foldable IOL and to assess biometric and anatomical factors associated with prediction error (PE).

METHODS

This post-hoc comparative formula analysis of a prospective interventional case series (24-month follow-up) included adult patients undergoing secondary implantation of a foldable single-piece IOL. IOL power was calculated using the SRK/T, Hoffer Q, Holladay 1, Barrett Universal II, and Kane formulas. Primary outcomes included PE, mean absolute error (MAE), and the proportion of eyes within PE of ±0.50 and ±1.00 diopters (D). Subgroup analysis using anterior segment optical coherence tomography evaluated associations between PE, IOL tilt, and decentration. Main outcome measures were MAE, proportion of eyes within PE ±0.50 and ±1.00 D, and correlation between PE and anatomical parameters (axial length, tilt, decentration).

RESULTS

A total of 226 eyes were included. MAE was lowest with Barrett Universal II (0.44 ± 0.37 D) and Kane (0.46 ± 0.39 D) (P < .0001, analysis of variance). The proportion of eyes with PE within ±0.50 D was 69.0% (129 of 187) for Barrett Universal II, 66.3% (124 of 187) for Kane, 53.5% (121 of 226) for SRK/T, 52.2% (118 of 226) for Hoffer Q, and 51.3% (116 of 226) for Holladay 1 (P < .0001, chi-square test). The proportion within ±1.00 D was 90.9% (170 of 187), 90.9% (170 of 187), 83.6% (189 of 226), 83.2% (188 of 226), and 82.3% (186 of 226), respectively (P < .0001, chi-square test). PE >±1.00 D was significantly associated with axial length <22 mm (myopic shift) or >26 mm (hyperopic shift) across all formulas (P = .022). In 86 eyes, IOL tilt showed a positive correlation with PE for Kane (B = 0.136, P < .001) and a weak negative correlation for Barrett Universal II (B = -0.062, P < .001), whereas no significant association was found for decentration (P > .05).

CONCLUSIONS

The Barrett Universal II and Kane formulas showed significantly better refractive prediction accuracy than third-generation formulas in eyes undergoing secondary sutureless intrascleral fixation of a one-piece foldable IOL. Axial length variability and IOL tilt significantly impacted PE across all formulas, underscoring the importance of biometric and anatomical factors in surgical planning.