目的：比较测得后三种不同角膜仪测量的角膜曲率，以评估其测量值之间的一致性。

方法：前瞻性研究。252例患者(252眼)使用IOL Master(IM)，Bausch & Lomb手动角膜仪(Man)以及TOPCON KR-8800自动角膜仪(Top)进行角膜曲率测量。记录并对比平均角膜曲率值。使用Bland Altman统计方法进行仪器间的一致性分析。

结果：1)IOL Master 和 手动角膜仪：IOL Master平均角膜曲率为44.62±1.52 D，手动角膜仪为44.60±1.52 D。 t-test显示差异具有统计学意义(P=0.001)； Bland-Altman图显示两种仪器间95%一致性区间(LOAs)为-0.22～0.22； 2)IOL Master 和自动角膜仪：IOL Master平均角膜曲率为44.62±1.52 D，自动角膜仪为44.46±1.53 D。t-test显示差异具有统计学意义(P<0.0001)。Bland-Altman图显示两种仪器间95%LOAs为 -0.24～0.55； 3)自动角膜仪和手动角膜仪：自动角膜仪平均角膜曲率为44.60±1.52 D，手动角膜仪为44.46±1.53 D。t-test显示差异具有统计学意义(P<0.0001)。Bland-Altman图显示两种仪器间95%LOAs为-0.30～0.57。

结论：使用不同的仪器获得的角膜曲率数据是不可替换的，这对于白内障外科医生在外科手术计划和结果评估方面具有重要意义。

METHODS: In this prospective study, keratometry was performed using an IOL Master(IM), a Bausch & Lomb manual keratometer(Man)and TOPCON KR-8800 autokeratometer(Top)on 252 eyes of patients recruited from camps for cataract surgery. The average keratometry values were recorded and compared. The agreements between the instruments were analyzed using the Bland-Altman statistical method. The main outcome measure was average keratometry values.

RESULTS:1)IOL Master and Manual keratometer: the mean corneal power was 44.62±1.52 D with the IOL Master and 44.60±1.52 D with the Manual keratometer. The paired t-test demonstrated a statistically significant difference in the mean corneal power between the IOL Master and Manual keratometer(P=0.001). The 95% LOAs of the two devices were -0.22 to 0.22 as shown in the Bland-Altman plot; 2)IOL Master and autokeratometer: the mean corneal power was 44.62±1.52 D with the IOL Master and 44.46±1.53 D with the autokeratometer. The paired t-test demonstrated a statistically significant difference in the mean corneal power between the IOL Master and autokeratometer(P<0.0001). The 95% LOAs of the two devices were -0.24 to 0.55 as shown in the Bland-Altman plot; 3)Autokeratometer and Manual keratometer: the mean corneal power was 44.60±1.52 D with Manual keratometer and 44.46±1.53 D with the autokeratometer. The paired ttest demonstrated a statistically significant difference in the mean corneal power between the autokeratometer and Manual keratometer(P<0.0001). The 95% LOAs of the two devices were -0.30 to 0.57 as shown in the Bland-Altman plot.

CONCLUSION: Keratometry data obtained with different instruments may not be interchangeable, a fact that has important implications for cataract surgeons with respect to both surgical planning and outcomes auditing.