角膜曲率计
数学
标准差
均方预测误差
统计
眼科
医学
视力
作者
Sohee Jeon,Leonardo Taroni,Enrico Lupardi,Kenneth J. Hoffer,Luigi Fontana,Domenico Schiano-Lomoriello,Hyunggoo Kwon,Giacomo Savini
出处
期刊:Journal of Refractive Surgery
[SLACK, Inc.]
日期:2023-03-01
卷期号:39 (3): 158-164
被引量:8
标识
DOI:10.3928/1081597x-20221221-03
摘要
To evaluate the accuracy of nine formulas to calculate the power of a new extended depth-of-focus intraocular lens (EDOF IOL), the AcrySof IQ Vivity (Alcon Laboratories, Inc), using measurements from two optical biometers, the IOLMaster 700 (Carl Zeiss Meditec AG) and Anterion (Heidelberg Engineering GmbH).After constant optimization, the accuracy of these formulas was analyzed in 101 eyes: Barrett Universal II, EVO 2.0, Haigis, Hoffer Q, Holladay 1, Kane, Olsen, RBF 3.0, and SRK/T. Both standard and total keratometry from the IOLMaster 700 and standard keratometry from the Anterion were used for each formula.Constant optimization provided slightly different values for the A-constant, which ranged between 118.99 and 119.16, depending on the formula and the optical biometer. According to the heteroscedastic test, within each keratometry modality the standard deviation of the SRK/T was significantly higher compared to that of the Holladay 1, Kane, Olsen, and RBF 3.0 formulas. The SRK/T formula provided less accurate results also when the absolute prediction errors were compared by Friedman test. According to McNemar's test with Holm corrections, statistically significant differences were found within each keratometry modality between the percentage of eyes with a prediction error within ±0.25 diopters obtained with the Olsen formula compared to the Holladay 1 and Hoffer Q formulas.Constant optimization remains a mandatory step to achieve the best outcomes with the new EDOF IOL: the same constant should not be used for all formulas and for both optical biometers. Different statistical tests revealed that older IOL formulas have lower accuracy compared to newer formulas. [J Refract Surg. 2023;39(3):158-164.].
科研通智能强力驱动
Strongly Powered by AbleSci AI