Batkov E N, Pashtayev N P, Mikhaylova V I
Cheboksary branch of S. Fyodorov Eye Microsurgery Federal State Institution, 10 Traktorostroiteley Pr., Cheboksary, Russian Federeation, 428027.
Vestn Oftalmol. 2019;135(1):21-27. doi: 10.17116/oftalma201913501121.
Accuracy of calculation of the intraocular lens (IOL) power in eyes with short axial length is inferior to one in emmetropic eyes. Most studies focus on relatively standard eyes.
To assess the accuracy of power calculation for IOL used to correct extreme hyperopia and to compare available formulas based on their predictive capacity.
Results of 13 implantations involving IOLs of at least 40 Diopters (D) in power were retrospectively evaluated. IOL power was calculated using five formulas: Haigis, Hoffer Q, HolladayI, SRKII, SRK/T. Mean numerical refractive prediction error (RPE) and mean absolute refractive prediction error (ARPE) were calculated. Mean and median ARPE were computed after optimizing the A constant. Proportions of eyes within certain RPE limits were compared between the formulas.
Mean RPE ranged from 1.43 to 11.71 D before adjustment and from 1.08 to 5.34 D after adjustment (p<0.0001). Haigis formula produced the least RPE, and SRKII - the most. Pairwise comparison by mean ARPE after adjustment revealed no statistically significant difference between Haigis and Hoffer Q formulas. Comparison of formulas by percentage of eyes with minimal RPE identified Haigis and Hoffer Q as the most accurate, while the difference between the two was not statistically significant. The difference between the most accurate formulas (Haigis and Hoffer Q) and the least accurate (SRKII) was statistically significant.
In eyes with extremely short anterior-posterior axis, prediction errors in IOL power calculations are relatively frequent (only 31-46% of eyes are within ±0.5 D) and warrant reduction. Among the evaluated formulas, Haigis and Hoffer Q are the most accurate. In order to improve the accuracy of IOL power calculations, it is necessary to employ personalized constants.
眼轴短的眼睛中人工晶状体（IOL）屈光度计算的准确性低于正视眼。大多数研究聚焦于相对标准的眼睛。
评估用于矫正极端远视的IOL屈光度计算的准确性，并根据预测能力比较可用公式。
回顾性评估13例植入至少40屈光度（D）IOL的手术结果。使用五个公式计算IOL屈光度：Haigis、Hoffer Q、HolladayI、SRKII、SRK/T。计算平均数值屈光预测误差（RPE）和平均绝对屈光预测误差（ARPE）。优化A常数后计算平均和中位数ARPE。比较各公式在特定RPE范围内的眼比例。
调整前平均RPE范围为1.43至11.71 D，调整后为1.08至5.34 D（p<0.0001）。Haigis公式产生的RPE最小，SRKII公式产生的RPE最大。调整后按平均ARPE进行两两比较，Haigis和Hoffer Q公式之间无统计学显著差异。按最小RPE的眼百分比比较公式，发现Haigis和Hoffer Q最准确，两者之间差异无统计学意义。最准确公式（Haigis和Hoffer Q）与最不准确公式（SRKII）之间的差异有统计学意义。
在眼前后轴极短的眼睛中，IOL屈光度计算中的预测误差相对常见（仅31 - 46%的眼睛在±0.5 D范围内），需要减少。在所评估的公式中，Haigis和Hoffer Q最准确。为提高IOL屈光度计算的准确性，有必要采用个性化常数。
