Refractive Outcomes: IOL Formulas for Patients With Previous PPV
Journal of Cataract & Refractive Surgery
Published online Dec. 21, 2018
Although refractive surprises after phacoemulsification have become less common with newer formulas for calculating the power of IOLs, estimates can be inaccurate for some patients, including those with previous ocular surgery. Lamson et al. looked at refractive outcomes of cataract surgery in previously vitrectomized eyes and compared the accuracy of various formulas for calculating IOL power. They found that, regardless of the calculation method, refractive outcomes after cataract extraction in vitrectomized eyes were more variable and hyperopic than the predicted outcomes. Of the formulas used in their study, Holladay 2 provided the best estimates.
This retrospective study involved a record review of phacoemulsifications performed from 2013 to 2017 (61 eyes of 57 patients; mean age, 60 years). Patients with previous pars plana vitrectomy (PPV) in the same eye were included. However, patients with a history of refractive surgery or silicone oil in the eye—or with any other factor that could preclude accurate calculation of IOL power—were excluded from the study.
The mean postoperative spherical equivalent was –0.16 D. Mean prediction errors were as follows: 0.30 ± 0.82 D for Holladay 1; –0.09 ± 0.76 D for Wang/Koch adjusted (WKA) Holladay 1; 0.23 ± 0.76 D for Holladay 2; 0.25 ± 0.81 D for SRK/T; 0.04 ± 0.85 D for WKA SRK/T; 0.33 ± 0.80 D for Hill-Radial Basis Function; 0.45 ± 0.80 D for Ladas; and 0.30 ± 0.82 D for Barrett.
The formula with the highest percentage of predictions within ±0.50 D of the postoperative outcome was Holladay 2 (60.42%). Significant differences between the predicted and actual refractive outcomes were found with all methods (p < .05) except WKA Holladay and WKA SRK/T. Intraclass correlation showed low repeatability (<0.50) for all formulas.
The authors acknowledged that the study is limited by its retrospective nature and small size, which did not allow for subgroup analyses by axial length or other parameters.
Moreover, the indications for vitrectomy varied widely, and differences in retinal pathology may have resulted in anatomic differences that affected formula performance.
Predicting refractive outcomes for this population is challenging, and patients should be counseled accordingly. Larger studies are needed to determine the best methods for choosing IOLs for patients who have previously undergone eye surgery.
The original article can be found here.