Predicting pseudophakic refractive error: Interplay of biometry prediction error, anterior chamber depth, and changes...

To quantify pseudophakic refractive error prediction with the Hill-RBF, Barrett Universal II, and SRK/T formulas and to evaluate the temporal effects of anterior chamber depth (ACD) and keratometric changes on postoperative refraction.

Department of Ophthalmology, University of Auckland and Auckland District Health Board, Auckland, New Zealand.

Prospective case series.

Patients listed for cataract surgery were prospectively recruited. Optimum intraocular lens (IOL) power and predicted refractive outcomes were calculated, and the predicted refraction was compared with objective refractive outcomes at 1-week and 1-month and 3-month follow-ups.

The study comprised 100 patients (100 eyes). The mean axial length was 23.4 mm ± 1.1 (SD). The mean keratometry was 43.9 ± 1.3 diopters (D). The mean absolute prediction errors at 3-months were: Universal II: 0.50 ± 0.46 D, RBF: 0.49 ± 0.43 D, and SRK/T: 0.52 ± 0.49 D. The Universal II formula was better than all other methods at 1 week (P = .02); no method was superior at 1 month (P = .46) or 3 months (P = .37). The RBF method predicted the highest proportion of eyes within ±0.25 D at 3 months. Keratometric changes primarily occurred 1 to 4 weeks postoperatively (P = .03) and trended toward corneal steepening (+0.11 D, P < .01). A significant mean hyperopic shift occurred 4 to 12 weeks postoperatively (+0.18 D, P < .01), accompanied by a trend toward increasing ACD (+0.04 mm, P = .03).

The Universal II and RBF methods were better than SRK/T in medium length eyes (22 to 24.5 mm). It might be prudent to expect a progressive hyperopic shift in the 3 months after surgery because of posterior IOL migration.