To investigate the difference between the segmented axial length (AL) and the displayed AL on an optical low-coherence reflectometry (OLCR) biometer and to compare the refractive prediction errors calculated using the segmented and displayed ALs. Retrospective case series. Four thousand nine hundred ninety-two eyes from 4992 patients in the theoretical study and 1758 eyes from 1758 patients in the refractive prediction error comparison. First, we calculated the segmented AL as the sum of geometrical ocular segments converted from the optical path length (OPL) in each medium. To convert the OPL to a geometrical distance in each medium, we used 4 sets of group refractive indices. Then, the mean absolute prediction error (MAE) was calculated with the displayed AL and segmented AL using 6 intraocular lens power formulas: Olsen, Barrett Universal II (Barrett), Haigis, Hoffer Q, Holladay 1, and Sanders-Retzlaff-Kraff trial (SRK/T). Segmented AL, difference in AL (segmented AL minus displayed AL), MAE, and percentage of eyes within 0.5 diopter (D) of error. The segmented ALs were up to 0.29 mm longer in short eyes and 0.50 mm shorter in long eyes. The differences in ALs were correlated negatively with the displayed ALs (r values, -0.941 to -0.913; P < 0.001). The MAEs were significantly lower using segmented ALs for all formulas except the Olsen in both the entire group and the long eye subgroup (AL, ≥26 mm) and for the Holladay 1 and Hoffer Q in the short eye subgroup (AL, ≤ 22 mm). Use of segmented ALs produced a greater percentage of eyes within 0.5 D of error for all formulas except the Olsen and Haigis for the entire group, for long eyes, and for the Holladay 1 in short eyes. The segmented ALs were longer in short eyes and shorter in long eyes compared with the displayed ALs calculated with a single group refractive index for the entire eye. The refractive accuracy with segmented ALs was improved in short eyes with the Hoffer Q and Holladay 1 formulas and in long eyes with all formulas except the Olsen formula.
Read full abstract