OR294
   
 

Intraocular lens power calculation after laser refractive surgery: Corrective algorithm for corneal power estimation

Haike. Guo¹, Haiying. Jin1,2, MD MD, Gerd. U. Auffarth², MD

¹Guangdong Eye Institute, Guangdong Academy of Medical Sciences, Guangzhou, China.
²International Vision Correction Research Centre, University of Heidelberg, Germany.

Purpose: To develop and evaluate an algorithm for corneal power estimation in intraocular lens (IOL) power calculation after myopic laser refractive surgery using direct cornea measuring.

Methods: Corneal parameters of 206 eyes of Chinese refractive patients were investigated by rotating Scheimpflug camera (Pentacam). Corneal optical power (Koptical) calculated by Gaussian optics formula was simplified as Koptical=Kanteior +K2 (Kanterior =anterior corneal power; Kposterior=posterior corneal power; K2=Kposterior-Kanterior×Kposterior×corneal thickness/1.376). The variation and the change of K2 induced by refractive surgery were analyzed. A corrective algorithm to calculate Koptical using mean K2 (-6.15 diopters, D), Kcorrective=1.114×measured K-6.15, was derived based on statistical analysis, which was in accordance with modified Maloney method. IOL powers of 49 post-refractive cases were calculated using Kcorrective.

Results: Mean K2 of post-refractive corneas were -6.15±0.20 D and. Mean change of K2 induced by refractive surgery was -0.05±0.12 D. Low variations of K2 were observed: 95% confident intervals of K2 were -6.43- to -5.65 D (pre-refractive), and -6.44 to -5.78 D (post-refractive). Using Kcorrective in IOL power calculation for post-refractive cases, mean absolute prediction errors were 0.60±0.54 D (Haigis), 0.61±0.52 D (double-K Hoffer Q) and 0.59±0.49 D (double-K SRK T), respectively.

Conclusions: The algorithm that induces low error in corneal power estimation is a relatively reliable method in IOL power calculation for previous myopic laser refractive cases. The algorithm for corneal power estimation for Asian was similar to Caucasian, comparing to our previous research (Kcorrective=1.114×measured K-6.10).


 
RANZCO