目的：探讨Olsen公式对高度近视合并白内障患者术后屈光力的预测性。

方法：回顾性分析自2016-10-10/2019-08-20于我院行白内障超声乳化联合人工晶状体(IOL)植入术的高度近视合并白内障患者65例101眼，根据眼轴长度(AL)进行分组，A组(26mm30mm)22例31眼。采用IOL Master自带的SRK/T、Haigis和Olsen公式计算术后理论屈光度(即所植入IOL度数对应的各公式的预测屈光度)，并记录术后3mo实际屈光度，计算绝对屈光误差(MAE)。

结果：三组患者采用Olsen公式对应的术后3mo MAE分别为0.15(0.04，0.22)、0.19(0.14，0.27)、0.26(0.115，0.455)D，无明显差异(P>0.05)，且对于相同眼轴长度范围的患者，采用Olsen公式对应的术后3mo MAE最小。采用Bland-Altman法分析采用三种IOL屈光度计算公式对应的术后理论屈光度与术后3mo实际屈光度的一致性，结果表明，术后3mo实际屈光度与Olsen公式对应的术后理论屈光度最为接近，与SRK/T公式对应的术后理论屈光度一致性最差。

结论：Olsen公式对高度近视合并白内障患者术后屈光力的预测准确性和稳定性优于SRK/T、Haigis公式，且Olsen公式受眼轴长度影响最小。

METHODS: This retrospective study reviewed 65 patients(101 eyes)with high myopia and who had phacoemulsification combined with intraocular lens implantation from October 10^th 2016 to August 20^th 2019 in our hospital. They were divided into three groups according to the axis length: group A(26mm 30mm)22 cases(31 eyes). The postoperative theoretical diopter(that is, the preoperative predictive diopter of each formula corresponding to the degree of IOL implanted)of intraocular lens was calculated by SRK/T, Haigis and Olsen formulas of IOL-master. The actual postoperative 3mo diopter was recorded and the absolute refractive error(MAE)was calculated.

RESULTS: The postoperative 3mo MAE in the three groups calculated with Olsen formula was 0.15(0.04, 0.22), 0.19(0.14, 0.27), 0.26(0.115, 0.455)D respectively(P>0.05). For patients with the same axial length, the postoperative 3mo MAE of Olsen formula was the smallest. The Bland-Altman method was used to analyze the consistency between the postoperative theoretical diopter and the postoperative 3mo actual diopter. The results showed that the postoperative 3mo actual diopter was the closest to the postoperative theoretical diopter of the Olsen formula, while the postoperative theoretical diopter of the SRK/T formula was the worst.

CONCLUSION: For patients with high myopia and cataract, the accuracy and stability of Olsen formula is better than that of SRK/T and Haigis formula, and Olsen formula is least affected by eye axis length.