Purpose: 

To evaluate the prediction error (PE) variance and absolute median PE of different intraocular lens (IOL) calculation formulas including last-generation formulas such as Barrett True-K with K, Okulix and total keratometry (TK)-based calculations with Haigis, and Barrett True-K in a simulation model in post–small-incision lenticule extraction (SMILE) eyes.

Settings: 

Department of Ophthalmology, University Hospital Marburg, Marburg, Germany.

Design: 

Prospective study.

Methods: 

Preoperative measurements included IOL power calculation before and after SMILE surgery. The target refraction was set to be the lowest myopic refractive error in pre-SMILE eyes. The IOL power targeting at the lowest myopic refractive error in pre-SMILE eyes was selected for the post-SMILE IOL calculation of the same eye. The difference between the predicted refraction of pre- and post-SMILE eyes with the same IOL power was defined as IOL difference. The refractive change induced by SMILE was defined as the difference between preoperative and postoperative manifest refraction.

Results: 

98 eyes from 49 patients underwent bilateral myopic SMILE. The PE variance of Okulix was not significantly different compared with Barrett True-K with TK (P = .471). The SDs of the mean PEs were ±0.413 D (Haigis-TK), ±0.453 D (Okulix), ±0.471 D (Barrett True-K with TK), ±0.556 D (Haigis-L), and ±0.576 D (Barrett True-K with K). The mean absolute PE was 0.340 D, 0.353 D, 0.404 D, 0.511 D, and 0.715 D for Haigis-TK, Okulix, Barrett True-K with TK, Barrett True-K with K, and Haigis-L, respectively. The highest percentage of eyes within ±0.50 D was achieved by Okulix, followed by Haigis-TK, Barrett True-K with TK, Barrett True-K with K, and Haigis-L.

Conclusions: 

Results suggest that Haigis in combination with TK, Okulix, and Barrett True-K with and without TK offer good options for accurate IOL power calculation after SMILE. Haigis-L showed a tendency for myopic shift in eyes after previous SMILE.