Abstract
Purpose: :
To develop a systematic method for quantifying pre-treatment adjustments to the ablation sphere in patients undergoing wavefront LASIK.
Methods: :
Data from 356 consecutive cases of CustomVue LASIK (<10D sphere; <3D cyl) performed in 2008 using Wavescan 3.67 software and an AMO S4 IR excimer laser (AMO, Santa Clara, CA) were tabulated in Minitab 15 (www.minitab.com). Multiple regression modeling (best subsets and stepwise) was then used to derive a weighted relationship between the achieved ablation and a variety of possible influential factors including age, central corneal pachymetry, 4mm pupil wavefront refraction spherical equivalent (WRSE), manifest refraction spherical equivalent (MRSE) minus 4mm pupil WRSE, absolute cylinder, and maximum pupil minus 3mm pupil WRSE. Spherical equivalent (attempted versus achieved) results from 192 similar consecutive cases of myopic CustomVue LASIK treated with the multiple regression derived nomogram in 2010 were then compared to those from the original 2008 data set.
Results: :
Factors found to influence the achieved refractive spherical equivalent change in regression modeling were (in weight order) 4mm pupil WRSE, MRSE minus 4mm pupil WRSE, maximum pupil minus 3mm pupil WRSE, and pachymetry. Variance in postoperative spherical equivalent results was halved in 2010 results using the new nomogram (0.10D in 2010 versus 0.21D in 2008; p<0.0001 Bartlett test). The (R2) measure of linear fit was increased from 0.958 in 2008 to 0.984 in 2010.
Conclusions: :
Multiple regression modeling can be used to take account of a variety of possible influences in enhancing the accuracy of wavefront LASIK. Nomograms developed through regression modeling can be used to derive personalized pre-treatment adjustments to the ablation sphere.
Keywords: cornea: clinical science • laser • myopia