周边前房深度扫描分析仪的可重复性及一致性研究——邯郸眼病研究

Repeatability and reproducibility of scanning peripheral anterior chamber depth analyzer——the Handan Eye Study

背景 如何早期发现原发性闭角型青光眼（PACG）患者或对其高危人群进行筛查是防治PACG的瓶颈问题.目前常用的评价前房深度（ACD）及房角形态的方法均有其不足.周边前房深度扫描分析仪（SPAC）能够测量ACD和评估房角关闭的危险性,具有操作简便、客观、非侵入性等特点. 目的 探讨SPAC测量前房相关指标及对可关闭房角判断的可重复性和不同测量者间的一致性. 方法 本研究为前瞻性横断面研究方法.于2013年2月在邯郸眼病研究5年随访调查中纳入年龄≥35岁、前房较浅、房角较窄者（改良的van Herick分级≤3且房角镜检查Shaffer分级≤2）30例30眼作为浅前房组,纳入年龄和性别匹配、ACD正常的成年人（改良的van Herick分级≥4）30人30眼作为正常对照组.受检者接受视力、验光、裂隙灯显微镜等常规眼科检查后,由两位测量者A和B在相同的测试条件下分别对每位受检者任一眼别连续进行3次SPAC测量,得到ACD、中央角膜厚度（CCT）、角膜曲率半径（CRC）和可关闭房角的判断情况（以SPAC数字分级≤5和/或定性分级为S、P为标准）等指标.ACD、CCT和CRC测量结果的可重复性用重复性限（RL）值进行评价,测量者间一致性用组内相关系数（ICC）进行评价,两位测量者间对可关闭房角判断的一致性采用Kappa值进行评价. 结果 测量者A对正常对照组和浅前房组测得的ACD、CCT和CRC的RL分别为0.36％ ～3.29％和0.90％ ～4.72％,测量者B对正常对照组和浅前房组测得的ACD、CCT和CRC的RL分别为0.38％ ～2.55％和0.51％ ～3.42％.两位测量者对正常对照组测得的ACD、CCT和CRC的ICC分别为0.966、0.859和0.989,对浅前房组测得的ACD、CCT和CRC的ICC分别为0.915、0.905和0.971;对可关闭房角判断结果的Kappa值分别为1.000和0.520. 结论 SPAC测量作为评价正常者和浅前房者ACD、CCT和CRC均有良好的可重复性和测量者间一致性;可关闭房角判断方面,SPAC在正常者具有良好的测量者间一致性,在浅前房者一致性为中度.

Background Early diagnosis of primary angle-closure glaucoma （PACG） and screening or case detection of high risk population have become a major challenge in the prevention and treatment of PACG.Commonly used methods for evaluation of anterior chamber depth （ACD） and anterior chamber angle nowadays have their shortcomings.Scanning peripheral anterior chamber depth analyzer （SPAC） measures ACD and evaluates the risk of angle closure.Also,SPAC is handy,objective and noninvasive.Objective This study was to evaluate the repeatability （within-rater differences） and reproducibility （between-rater differences） of SPAC to measure anterior chamber-related parameters and agreement of SPAC to estimate occludable angle between two operators.Methods A prospective cross-sectional study was designed.Thirty subjects with shallow ACD （modified van Herick grading≤ 3 and Shaffer grading≤2） aged 35 years and older,who met the inclusion criterion were enrolled as shallow anterior chamber group and 30 age-and gender-matched normal adults （modified van Herick grading≤3 and Shaffer grading≤2）from the follow-up research of the Handan Eye Study in February 2013.After an elementary eye examination,a randomly selected eye of each subject in both groups received three consecutive SPAC measurements by each of the two examiners,respectively to obtain ACD,central corneal thickness （CCT） and corneal radius of curvature （CRC）.Meanwhile,the identity of occludable angle as the numerical grade was 5 or fewer and/or the categorical grade being S or P was also obtained.The repeatability of each examiner was evaluated by the repeatability limit （RL） and the reproducibility of the two examiners was evaluated by intraclass correlation coefficient （ICC）.The agreement of occludable angle estimation between two operators was evaluated using Kappa coefficient.Results The RLs percentage for ACD,CCT and CRC ranged from 0.36％ to 3.29％,and 0.90％ to 4.72％ in the normal control group and shallow anterior chamber group from operator A,and those from operator B were 0.38％ to 2.55％,and 0.51％ to 3.42％.The ICCs for ACD,CCT and C RC of SPAC were 0.966,0.859,0.989 and 0.915,0.905,0.971,respectively from operator A and operator B.The Kappa coefficients for the identity of occludable angle between two operators were 1.000 and 0.520 in the normal control group and the shallow anterior chamber group,respectively.Conclusions Measurements of the three parameters provided by SPAC in normal and shallow anterior chamber subjects show a good repeatability and reproducibility.The results of identity of occludable angle from two examiners have a good agreement for normal subjects and a moderate reproducibility for subjects with shallow ACD.