目的：比较SW-9000与IOL Master 500两种光学生物测量仪测量眼轴长度(AL)、前房深度(ACD)、平坦轴角膜曲率(K1)和陡峭轴角膜曲率(K2)的差异性，并对两种仪器的测量结果进行一致性评价。

方法：前瞻性临床研究。选取2019-09/2020-07在我科屈光手术中心接受术前检查的屈光不正患者129例258眼，其中低中度近视(-0.75D≤等效球镜度<-6D)159眼，高度近视(等效球镜度≥-6D)99 眼，分别使用SW-9000和IOL Master 500进行AL、ACD、K1和K2的测量。两种仪器测量结果比较采用配对t检验，两种仪器测量结果的相关性采用 Pearson相关分析，两种仪器间测量数据的一致性采用Bland-Altman分析。

结果：SW-9000 和IOL Master 500在测量低中度近视患者的AL和ACD时均有差异(均P<0.01)，而测量K1与K2 时均无差异(均P>0.05)； SW-9000 和IOL Master 500测量高度近视患者的AL、ACD和K1值时均无差异(均P>0.05)，在测量K2 时有差异(P<0.05)。相关性分析结果显示，两种仪器测量低中度和高度近视患者的AL、ACD、K1和K2 时均呈高度相关(均P<0.01)。Bland-Altman一致性分析结果显示，两种方法测量低中度和高度近视患者的AL时一致性较好，而测量ACD、K1和K2的一致性较差。

结论：SW-9000与IOL Master 500在测量屈光不正患者AL时具有良好的一致性，对ACD和角膜曲率的测量，两者一致性较差，在使用中需结合临床实际谨慎选择。

METHODS: Totally 258 eyes of 129 patients with ametropia were measured with SW-9000 and IOL Master 500. Among them, 159 eyes were low to moderate myopia(-0.75D≤ spherical equivalent <-6D)and 99 eyes were high myopia(spherical equivalent ≥-6D). Comparison between two devices were performed for AL, ACD, K1, and K2. The difference between two instruments were evaluated with a paired samples t-test. The correlation between the measurements was evaluated using Pearson correlation coefficients. The agreement between the devices was evaluated with Bland-Altman plots.

RESULTS: There was significant difference between SW-9000 and IOL Master 500 in measuring AL and ACD(all P<0.01), but there was no significant difference between K1 and K2 in patients with low to moderate myopia. There was no significant difference in AL, ACD and K1 between SW-9000 and IOL Master 500(all P>0.05), but there was significant difference in K2(P<0.05)in patients with high myopia. Pearson correlation analysis showed that AL, ACD, K1 and K2 of patients with low to moderate and high myopia were highly correlated(all P<0.01). Bland-Altman consistency analysis showed that the two methods had good consistency in measuring AL in patients with low to moderate and high myopia, but poor consistency in measuring ACD, K1 and K2.

CONCLUSION: SW-9000 and IOL Master 500 have good consistency in measuring the AL of people with ametropia. The measurement of ACD and corneal curvature has poor consistency, and should be carefully selected in combination with clinical practice.