Photorefractive keratectomy for hyperopia - Long-term nonlinear and vectoranalysis of refractive outcome

Purpose: To characterize the refractive changes after excimer laser photore fractive keratectomy for the correction of hyperopia over a follow-up up to 3 years and to assess refractive stability and changes in astigmatism. Design: Noncomparative, nonrandomized, retrospective, interventional case s eries. Participants: Thirty-eight hyperopic eyes of 28 patients (age range, 33-62 years) with refraction in the range +1.00 to +8.00 diopters (D). Mean attem pted correction was +3.33 +/- 0.98 D (range, +1.00 to +4.00 D). Data were c ompared to those from 216 eyes treated for myopia in the range -1.00 to -12 .70 D. Intervention: The hyperopic correction was made using an erodible mask inse rted in the laser optical pathway to produce a circular ablation measuring 6.5 mm in diameter. An axicon was then used to create a blend transition zo ne from 6.5 mm up to 9.4 mm in diameter. Eyes were evaluated 3 to 11 times (5.5 +/- 2.4) over a 3- to 34-month follow-up (16.8 +/- 8.4 months). Main Outcome Measures: Vector analysis of refractive error, applying a nonl inear statistical model fitting the spherical equivalent, and the sphere co mponent data. The fit parameters were the long-term error at stabilization (epsilon(infinity)), the amount of regression (epsilon(0)), being the diffe rence of refractive errors immediately after surgery and at stabilization, and the time constant (T-1/2) giving the temporal scale length by which the overcorrection halves (regression half-life). Cylinder was analyzed by a l inear regression. Results: The initial overcorrection was much larger after hyperopic treatme nts than myopic ones (epsilon(0) = -3.26 +/- 0.35 D vs. +1.43 +/- 0.35 D), and it takes typically four times longer to regress (T-1/2 = 3.30 +/- 0.91 months). Refractive stabilization is reached after more than 1 year, with a satisfactory refractive result. The hyperopic treatment induces a mean ast igmatism of 1.00 D, which remains constant throughout the follow-up, and te nds to be aligned along the with-the-rule meridian. Conclusions: The advantages of a reasonably well-designed algorithm to corr ect hyperopia (epsilon(infinity) = +0.20 +/- 0.23 D) are counterbalanced by the long time to refractive stabilization and by the induced astigmatism.