Comparison of Intraocular Lens Power Calculation Formulae in Pediatric Eyes

Purpose To evaluate accuracy of intraocular lens (IOL) power calculation formulae (SRK II, SRK/T, Holladay 1, Hoffer Q) in pediatric eyes. Design Retrospective case series. Participants One hundred thirty-five eyes of 96 children with congenital, developmental, or acquired cataracts who underwent uncomplicated cataract surgery and IOL implantation by a single surgeon over a 10-year period. Methods Axial length (AL), keratometry (K), and manufacturer's A constant were employed in 4 common IOL power calculation formulae to predict the refractive outcome. Retinoscopy was measured at 4 to 8 weeks postoperatively and converted to spherical equivalent. For analysis, eyes were grouped by age at surgery, AL, and mean K. Main Outcome Measures We determined the prediction error (PE) = predicted refraction − actual refraction and the absolute PE = | predicted refraction − actual refraction |. The formula that gave the best prediction (minimum PE) was determined. Results The mean age at surgery was 6.4 years. Mean absolute PE was 1.11 for the SRK II, 0.84 for SRK/T, 0.76 for Holladay, and 0.76 for Hoffer Q formulae. There was a trend toward greater PE in eyes of younger children (≤2 years), shorter AL (AL ≤ 22 mm) and steeper corneas (mean K > 43.5 diopters [D]). On comparing absolute PE obtained with 4 formulae in each patient, Hoffer Q gave the minimum PE in 46% of eyes compared with 23% with SRK II, 18.5% with SRK/T, and 12.5% with Holladay 1. The SRK/T, Holladay 1, and Hoffer Q were similar in accurately predicting refractive error within ±0.5 D in about 43% eyes. When clinically significant deviation in PE occurred (>0.5 D), there was usually an undercorrection (72%), except for Hoffer Q, which was almost as likely to overcorrect as undercorrect (44% vs 56%). The PE was lower with office measurements when compared with anesthesia measurements, owing probably to better fixation in older children with higher ALs. Conclusion The PE was insignificant (PE ≤ 0.5 D) in 43% eyes, and similar for all formulae. However, the Hoffer Q was predictable for the highest number of eyes. When the PE was >0.5 D, most formulae gave an undercorrection, except for the Hoffer Q, which the surgeon may want to consider when targeting postoperative refractions. Financial Disclosure(s) The authors have no proprietary or commercial interest in any of the materials discussed in this article.