Comparison of changes in manifest refraction and corneal power after photorefractive keratectomy

PURPOSE: To determine which corneal curvature values most closely correlate to change in manifest refraction after excimer laser photorefractive keratectomy. METHODS: In a prospective study at the Cullen Eye Institute, excimer laser photorefractive keratectomy was performed on 21 eyes of 21 patients (mean age, 38.01 +/- 6.65 years). Preoperative refractive errors ranged from -2.25 diopters to -8.75 diopters (mean, -5.14 +/- 2.09 diopters). Preoperatively and 1 month postoperatively, we determined the spherical equivalent of the subjective manifest refraction (corrected for a 12-mm vertex distance) and measured corneal power using standard keratometry (Bausch & Lomb Keratometer; Rochester, New York) and computerized videokeratography (EyeSys Corneal Analysis System; Premier Laser Systems Inc, Houston, Texas). We collected 15 corneal values: standard keratometry and 14 computerized videokeratography values calculated using the axial, instantaneous, and refractive formulas. All calculations were performed with 1.3375 and 1.376 for the refractive index of the cornea, For each of the corneal values, we subtracted the change in corneal power from the change in manifest refraction and calculated for this difference the means, SDs, correlations, and regressions. RESULTS: Mean differences between change in refraction and change in corneal power were lower when for a refractive index of 1.376 than for 1.3375, were lowest for the most central measurement points, and displayed a high SD. A value of 1.408 for the refractive index would be required to optimize the correlation between change in manifest refraction and effective refractive power of the central 3 mm of the cornea. CONCLUSIONS: For individual patients who have undergone photorefractive keratectomy, changes in corneal values determined by computerized videokeratography or by standard keratometry do not reliably predict change in manifest refraction. (Am J Ophthalmol 2000;129: 68-75. (C) 2000 by Elsevier Science Inc. All rights reserved.)