Accuracy of IOL power calculation using the last-generation Olsen formula: a comparison with commonly used IOL formulas

Session Details

Session Title: Presented Poster Session 15: Cataract Surgery Outcomes 3

Session Date/Time: Monday 15/09/2014 | 09:30-11:00

Paper Time: 09:45

Venue: Pod 3 (Poster Village)

First Author: : E.Toker TURKEY

Co Author(s): :    Ü. Taka Aydın   A. Özçelik Köse           

Abstract Details

Purpose:

To evaluate the accuracy of IOL power calculations using the last generation Olsen formula and to compare it with the current IOL power calculation formulas.

Setting:

Marmara University Medical School, Department of Ophthalmology, İstanbul, Turkey.

Methods:

One hundred and ninety eyes of 123 patients were included to this study. Biometric measurements were performed with the optical low-coherence reflectometry (Lenstar LS900, Haag-Streit AG). All patients underwent uncomplicated phacoemulsification by a single surgeon (E.T.) using a single IOL model (Acriva UD 613, VSY Biotechnology). The power of the implanted IOL was used to calculate the predicted postoperative spherical equivalent by five formulas: the Olsen, SRK/T, Holladay, Haigis and Hoffer Q. The mean and median absolute errors (MAE, MedAE) and percentage of eyes within ±0.25 D, ±0.50 D of predicted refraction and deviating over ±1.00 D of predicted refraction were calculated for each formula. The results were also classified into 3 groups based on axial length (less than 22 mm, in between 22-25 mm and more than 25 mm).

Results:

In all eyes, the MAE (MedAE) was smallest with the Olsen formula [0.32(0.27)] but the difference was not statistically significant when compared to the MAE(MedAE)s obtained by other formulas [SRK/T, 0.35(0.27); Holladay, 0.36(0.29); Haigis, 0.34(0.30); Hoffer Q, 0.36(0.28), p>0.05]. The percentage of eyes within ±0.25 D and ±0.50 D of predicted refraction was greatest in the Olsen formula (47.9%, 79.4%), followed by those in the Haigis (45.8%, 77.3%), SRK/T (44.7%, 76.8%), Holladay (44.2%, 75.8%), and Hoffer Q (43.7%, 74.2%). The proportion with a MAE deviating over ±1.00 D was lowest in the Olsen formula (2.1%) and ranged between 3.7-5.7% for other formulas. In the average axial length (AL) group, the Olsen formula had higher prediction accuracy compared to SRK/T (p=0.04), Haigis (p=0.02) and Hoffer Q (p=0.06) formulas. In the short AL group, the MAE generated by the Olsen formula was significantly higher than that generated by the SRK/T (0.16 D vs. 0.43 D, respectively, p=0.012) but comparable to those obtained by other formulas. In the long AL group, there was no statistically significant difference in the MAE (MedAE)s of all formulas.

Conclusions:

These results suggest that the Olsen formula performs well in eyes with average and long axial lengths yielding better or comparable refractive results compared with the other current formulas.

Financial Interest:

NONE