目的：通过比较患者难测眼的坐位、卧位A超和晶星900(Lenstar，LS900)三种方法测量眼轴长度的差异，为更准确进行难测眼眼轴测量提供指导。

方法：临床病例对照研究。对2019-05/2020-09在郑州市第二人民医院就诊的合并硅油眼、玻璃体积血或视网膜脱离的白内障患者及晶状体或人工晶状体脱位的患者102例102眼分别行LS900及坐位、卧位A超检查测量眼轴长度，并对三种方法测量结果进行统计学分析。

结果：LS900与A超眼轴测量检出率分别为83%和100%； A超坐位和卧位及LS900三种方法测量眼轴长度整体、硅油组、晶状体脱位组、玻璃体积血组比较均有差异(P<0.001)，整体卧位、硅油组卧位、晶状体脱位组卧位、玻璃体积血组卧位与LS900测量值眼轴长度均值比较有差异(均P<0.05)，而整体坐位和各难测眼组坐位与LS900测量眼轴长度比较均无差异； 三种测量方法所测值在95%一致性范围内一致性良好，A超坐位测量结果更接近LS900。

结论：在难测眼A超眼轴测量中，特别是眼部组织结构发生改变的疾病，改变常规卧位为坐位可提高测量结果的准确性，为临床医生制定治疗方案提供更可信的指导。

AIM: To provide guidance for more accurate measurement of axial length(AL)of difficult measuring eyes by comparing the differences in the AL of the patient's difficult measuring eye with three methods of the sitting position, supine position A-scan and Lenstar 900(Lenstar, LS900).

METHODS: Clinical case-control study. We selected 102 cases(102 eyes)including cataract patients with combined silicone oil filled, vitreous hemorrhage or retinal detachment and patients with dislocation of the lens or IOL in Zhengzhou Second Hospital from May 2019 to September 2020. AL were measured using LS900 and A-scan on sitting position and supine position respectively, and the results of the three methods were statistically analyzed.

RESULTS: The detection rates of LS900 and A-scan axial measurement were 83% and 100% respectively. Three methods of A scan in sitting position, supine position and LS900 to measure the overall AL, silicone oil group, lens dislocation group and vitreous hemorrhage group, the differences were statistically significant(P<0.001), The mean values of AL measured by overall A-scan, supine positions of silicone oil group, supine position of lens dislocation group and vitreous hemorrhage group were statistically significant differences with LS900 measurement(all P<0.05), while there was no statistical difference between the results of the overall sitting position and the difficult measuring eye groups' sitting position compared with the LS900 measurement of AL. The three measurements showed good consistency within the 95% consistency range, but the result of A-scan on sitting position was closer to LS900.

CONCLUSION: Changing the conventional decubitus position to the sitting position can improve the accuracy of the measurement results and provide clinicians with more reliable guidance for the treatment of patients with difficult measuring eyes of A-scan axial measurement, especially in diseases with altered ocular structure.