Purchase this article with an account.
Carrie Wright, Shruti Sudhakar, Hans Andrews, Tara L O’Rourke, Ingrid U Scott, Seth Pantanelli; Hill-RBF Calculator versus Holladay2 and SRK/T Formulas for Intraocular Lens Power Selection. Invest. Ophthalmol. Vis. Sci. 2017;58(8):1149.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To compare the Hill-RBF calculator to the Holladay2 and SRK/T formulas with respect to predicting residual refractive error in eyes with normal axial lengths.
The study was approved by the local Institutional Review Board. This retrospective consecutive case series included 31 eyes with axial lengths 22.0-25.2 mm that underwent cataract surgery with intraocular lens (IOL) implantation by a single surgeon between March and July 2016 at an academic medical center. All eyes underwent preoperative biometry measurements with IOLMaster 700 and received Bausch & Lomb MX60 one-piece IOL implants. For each eye, residual refractive error was predicted preoperatively using Holladay2 and SRK/T formulas, as well as the Hill-RBF calculator, with a physician-optimized lens constant of 119.4. Refraction was performed one month postoperatively. The mean numerical error for each formula was calculated as the difference between the predicted and actual postoperative spherical equivalent refractive error.
The mean numerical errors for the Holladay2, SRK/T, and Hill-RBF formulas were -0.07, -0.02, and 0.12, respectively (ANOVA p-value = <0.00003). Median numerical errors for the same three equations were 0.045, 0.035, and 0.16, respectively. Post-hoc pairwise comparisons of mean numerical error revealed that both the Holladay2 (p < 0.000007) and SRK/T (p = 0.0017) formulas performed significantly better than the Hill-RBF formula for the tested dataset, but were equivalent to each other (p = 0.24). The proportion of patients with a postoperative refractive error within 0.5D of predicted was 87.1% for Holladay 2, 77.4% for SRK/T, and 77.4% for Hill-RBF formulas.
Postoperative refractive error was within 0.5 D of predicted in greater than 75% of eyes for each of the three formulas. With respect to the mean numerical error, both the Holladay2 and SRK/T formulas performed better than the Hill-RBF, but further investigation using larger datasets are warranted.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.
This PDF is available to Subscribers Only