Abstract
Purpose :
Radial keratotomy (RK) incisions alter corneal curvature and lead to errors in predicting intraocular lens (IOL) power for cataract surgery. The purpose of this study is to compare prediction error of the Barrett True K No History (Barrett TKNH) formula and intraoperative aberrometry (IA) in eyes with prior RK.
Methods :
A retrospective study was conducted using records of all patients with prior RK at our institution who underwent cataract surgery with IA from 2014 to 2019. Refraction prediction error for IA and Barrett TKNH were compared. One sample t-tests were used to assess whether mean prediction errors were significantly different than zero. General linear modeling accounting for the correlation between eyes was used to determine whether absolute refraction prediction error differed between Barrett TKNH and IA.
Results :
A total of 48 eyes (32 patients) were included in this study. The median absolute refraction prediction errors for Barrett TKNH and IA were 0.50D and 0.49D respectively (Table 1A). The mean numerical errors were -0.12 ± 1.08D and -0.10 ± 1.08D, respectively (Figure 1), neither was significantly different than zero (p=0.43, p=0.53). General linear modeling comparing the absolute refraction prediction error for Barrett TKNH versus IA showed no significant difference (p=0.71). The refractive outcome fell within ±0.50 D of prediction for 50% of eyes with Barrett TKNH and 54.2% with IA, and 79.2% were within ±1.00 D of prediction for both techniques (Table 1B). Mean absolute prediction error increased with increasing number of RK cuts (grouped into <8 cuts, 8 cuts, >8 cuts) for both Barrett TKNH and IA (0.35D, 0.82 D, and 0.92D respectively for Barrett TKNH, and 0.30D, 0.87D, 0.94D respectively for IA), but groups were too small to draw statistical conclusions.
Conclusions :
There is no statistically significant difference between Barrett TKNH and intraoperative aberrometry in predicting postoperative refractive error in eyes with prior RK. Both Barrett TKNH and IA are reasonable methods of calculating IOL power for eyes with prior RK, however there is room for improvement as outcomes still lag behind those of other post-refractive, and non-post-refractive, eyes.
This is a 2020 ARVO Annual Meeting abstract.