目的：观察并分析免散瞳眼底照相在糖尿病视网膜病变(DR)筛查中的作用，为临床筛查提供依据。

方法：选取我院2016-12/2017-11确诊为糖尿病(DM)患者120例240眼为观察对象，由同一操作者逐一进行免散瞳眼底照相、散瞳后7方位眼底照相及眼底荧光血管造影检查(FFA)。以糖尿病视网膜病变国际临床分级为标准，采取盲法由同一医师分别对以上三项检查进行阅片、分级并记录。最后以眼底荧光血管造影检查为金标准，分别对前两种方法与眼底荧光造影的结果进行对比性研究，计算每种方法对不同程度的DR的敏感度、特异度、Youden指数、Kappa值，筛选出简单易行、有效的糖尿病视网膜病变的筛查方法。

结果：通过眼底血管造影筛查出糖尿病视网膜病变168眼(70.0%)，散瞳后眼底照相筛查出糖尿病视网膜病变160眼(66.7%)，免散瞳眼底照相筛查出糖尿病视网膜病变156眼(65.0%)。不同方法筛查出的糖尿病视网膜病变分级结果基本一致，差异无统计学意义(P>0.05)。对不同程度糖尿病视网膜病变进行筛查时，免散瞳眼底照相组的敏感度和特异度分别为92.9%、90.3%，与散瞳后眼底照相组的结果无明显差异。与金标准组(FFA)比较，其Youden指数(83.14%)接近于1，结果可靠度高，Kappa=0.81，检验具有一致性。对中度非增殖期糖尿病视网膜病变进行筛查时，免散瞳眼底照相组的敏感度和特异度的分别是90.6%、95.5%，与散瞳后眼底照相结果无明显差异。与金标准组比较，免散瞳眼底照相的Youden指数为86.09%，结果可靠度高，Kappa=0.86，检验具有一致性。

结论：免散瞳眼底照相作为一种筛查糖尿病视网膜病变的简单、准确率高的可行性方法。此项检查简单易行，无风险，可培训专科技师进行多点操作，借助如今方便的网络传输图像至有经验的眼科医师阅片、诊断，方便快捷，让患者就近诊治，避免了来回奔波的辛劳，对社会具有积极的意义。

METHODS:In our hospital from December 2016 to November 2017, 120 patients(240 eyes)was diagnosed as diabetes(DM), which were treated as the subjects of observation. By the same operator with non-mydriatic fundus photography, fundus photography and 7 range fundus fluorescein angiography(FFA)after mydriasis were taken. Taking the international clinical classification of diabetic retinopathy(DR)as the standard, the above three examinations were review, grade and record by the same physician by blind method. The fundus fluorescein angiography as the gold standard, the other two results were compared to detect the sensitivity, specificity, Youden index, Kappa value of the two for DR with different grade.

RESULTS: There was 70.0% eyes diagnosed as diabetic retinopathy after screened by fundus angiography, 66.7% by post-mydriatic fundus photography, 65.0% by non-mydriatic fundus photography. The grading results of diabetic retinopathy screened by different methods were basically consistent, with no significant difference(P>0.05). When screening for diabetic retinopathy of different degrees, the sensitivity and specificity of the non-mydriatic group were 92.9% and 90.3%, respectively. There was no significant difference between the results of the non-mydriatic group and the non-mydriatic group. Compared with the gold standard group(FFA), the Youden index(83.14%)was close to 1, with high reliability; Kappa=0.81, and the validation was consistent. When screening for moderately nonproliferative diabetic retinopathy, the sensitivity and specificity of non-mydriatic fundus photography were 90.6% and 95.5%; there was no significant difference between the results of non-mydriatic and the results of fundus photography after mydriasis. Compared to the gold standard group, the Youden index was 86.09%, the reliability is high, Kappa=0.86, and the test was consistency.

CONCLUSION:Non-mydriatic fundus photography can be used as a simple and accurate method for screening diabetic retinopathy. It is simple and easy to carry out without risk. It is easy to train specialist technicians for multi-point operation. With the help of today's convenient network, the image is transmitted to an experienced ophthalmologist for reading and diagnosis, which is convenient and fast, so that the patient can be diagnosed and treated nearby, which has positive significance for the society.