不同临床经验医师应用LOCSⅢ对晶状体混浊程度评分的一致性研究

Repeatability of Lens Opacity Classification System Ⅲ between physicians with different clinical experience

目的评价不同临床经验医师应用晶状体混浊分级系统(Lens Opacity Classification System Ⅲ,LOCSⅢ),基于裂隙灯及数码照片进行晶状体混浊程度评分的一致性。方法从邯郸眼病研究5 a随访项目中,随机选取2012年11月至2012年12月现场调查受试者51人(102眼)。首先由受过LOCSⅢ评分培训的高年资眼科副主任医师(评分员A)于现场参照LOCSⅢ标准图片,在裂隙灯下对晶状体核混浊程度(nuclear opacity,NO)、核颜色(nuclear color,NC)、皮质性白内障(cortical cataract,C)、后囊膜下白内障(posterior subcapsular cataract,PSC)进行评分,随后采集晶状体数码照片。评分员A和1名眼科住院医师(评分员B)根据晶状体数码照片独立进行NO、NC、C、PSC评分。然后一位经过LOCSⅢ晶状体评分培训的白内障专业副主任医生对评分员B进行培训,同时与评分员A进行讨论并统一对LOCSⅢ标准评分方法的理解后,评分员B进行第2次评分。均采用组内相关系数(intraclass correlation coefficient,ICC)分别检验评分员A基于裂隙灯及数码照片的LOCSⅢ晶状体评分的一致性,评分员B培训前、后的LOCSⅢ晶状体评分的一致性,以及评分员B培训前、后的LOCSⅢ晶状体评分与评分员A基于裂隙灯及数码照片的LOCSⅢ晶状体评分进行一致性检验。结果评分员A基于现场裂隙灯和晶状体数码照片的LOCSⅢ晶状体评分的ICC分别为:NO 0.89、NC 0.89、C 0.90、PSC0.65;评分员B培训前、后LOCSⅢ晶状体评分ICC分别为:NO 0.90、NC 0.90、C 0.89、PSC 0.12;评分员B培训前、后基于数码照片与评分员A基于裂隙灯的LOCSⅢ晶状体评分ICC分别为NO 0.77、0.82,NC 0.79、0.85,C 0.77、0.89,PSC 0.05、0.46;评分员B培训前、后与评分员A基于数码照片的LOCSⅢ晶状体评分比较ICC分别为:NO 0.83、0.81,NC 0.80、0.84,C 0.76、0.88,PSC 0.26、0.85。结论眼科临床经验对临床医师采用LOCSⅢ进行NO、NC、C评分影响不大,对PSC评分影响较大,而LOCSⅢ标准评分方法培训可以弥补眼科医师临床经验的不足,但不能提高基于裂隙灯和数码照片的LOCSⅢ对PSC评分的一致性。

Objective To evaluate the repeatability of Lens Opacity Classification SystemⅢ(LOCSⅢ)based on slit lamp and digital photos,as well as between different clinical physicians.Methods From November 2012 to December 2012,51 participants(102 eyes)were randomly selected from the five year follow-up study of Handan eye study.Firstly,a senior ophthalmologist(scorer A)graded lens nuclear opacities(NO),nuclear color(NC),cortical cataract(C)and posterior capsular cataract(PSC)under slit lamp on the scene,referring to the LOCSⅢstandard pictures.Then the lens digital photos were collected.The scorer A and one intern(scorer B)independently performed NO,NC,C and PSC scores according to the lens digital photographs.Then a cataract professional ophthalmologist,who was trained by the LOCSⅢlens score training,trained the scorer B.After scorer A and B discussed and unified the LOCSⅢstandard scoring method,scorer B scored the lens digital photographs for the second time.Repeatability of the LOCSⅢscore of the scorer A based on slit lamp and digital photograph was evaluated.Repeatability of the LOCSⅢscore of the scorer B before and after training was evaluated.The LOCSⅢscore of the scorer B before and after training was also compared to the LOCSⅢscore of the scorer A based on slit lamp and digital photos.Results Intraclass correlation coefficients(ICC)for the LOCSⅢscore of the scorer A based on digital photos and digital photos was 0.89 for NO and NC,0.90 for C and 0.65 for PSC.Intraclass correlation coefficients(ICC)for the LOCSⅢscore of the scorer B before and after training was 0.90 for NO and NC,0.89 for C and 0.12 for PSC.Intraclass correlation coefficients(ICC)for the LOCSⅢscore of the scorer B before training and the LOCSⅢscore of the scorer A based on slit lamp was 0.77 for NO,0.79 for NC,0.77 for C and 0.05 for PSC.After training,the ICC was 0.82 for NO,0.85 for NC,0.89 for C and 0.46 for PSC.Intraclass correlation coefficients(ICC)for the LOCSⅢscore of the scorer B after training and the LOCSⅢscore of the scorer A based on digital photos was 0.83 for NO,0.80 for NC,0.76 for C and 0.26 for PSC.After training,the ICC was 0.81 for NO,0.84 for NC,0.88 for C and 0.85 for PSC.Conclusion Clinical experience has little effect on the NO,NC,and C scores using LOCSⅢ,but has great influence on the PSC score.After LOCSⅢstandard training,the accuracy of scoring on PSC could be improved,but the repeatability of PSC score based on the slit lamp and the digital photos did not improve.