Mizrachi Yossi, Knyazer Boris, Guigui Sara, Rosen Shirley, Lifshitz Tova, Belfair Nadav, Klemperer Itamar, Schneck Marina, Levy Jaime
Faculty of Health Sciences, Ben-Gurion University of the Negev, Beersheba, Israel.
Int Ophthalmol. 2014 Aug;34(4):831-7. doi: 10.1007/s10792-013-9887-3. Epub 2013 Dec 1.
To evaluate the effectiveness of the non-mydriatic digital camera for diabetic retinopathy (DR) screening. Secondary purposes of the study were to characterize diabetic patients being screened for the presence of DR and to calculate the sensitivity, specificity, and positive predictive value of the test. All 6,962 consecutive patients with type 2 diabetes undergoing non-mydriatic digital retinal photography between January 1, 2009 and June 30, 2010 in eight community health clinics in the south of the country were included. Comparison of a random sample of patients who underwent non-mydriatic retinal photography, and who were also examined by an ophthalmologist with pupil dilation was also performed. The average age of all patients was 64.2 years. A total of 5,960 cases (85.6 % of all photographs) were of adequate quality for the diagnosis. DR of any degree was found in 1,092 (18.3 %) patients. Normal fundus pictures were found in 49.4 % of patients. In 32.2 % of cases, non-DR pathologies were found. Among cases in which DR was found, 73.3 % (801 cases) had mild non-proliferative retinopathy (NPDR), 7.1 % (77 cases) had moderate NPDR, 6.8 % (74 cases) had proliferative retinopathy, and 12.8 % (140 cases) had diabetic macular edema. Older patients had more chance of having poor quality pictures (p < 0.001 between patients older and younger than 70 years). When non-mydriatic fundus photography was compared with dilated fundus examination by an ophthalmologist, sensitivity of 99.3 %, specificity of 88.3 %, and positive predictive value of 85.3 % were found. Non-mydriatic digital retinal photography is an efficient method for DR screening. The test has high sensitivity and specificity. The test, as performed in community health centers in the south of the country, contributed to the early diagnosis of >1,000 cases of DR. Many patients can be followed up in a fast and efficient way, although the test cannot replace a complete eye examination after pupil dilation mainly in older people. Other non-DR sight-threatening ocular pathologies can be also detected.
评估免散瞳数码相机用于糖尿病视网膜病变（DR）筛查的有效性。该研究的次要目的是描述接受DR筛查的糖尿病患者的特征，并计算该检测的敏感性、特异性和阳性预测值。纳入了2009年1月1日至2010年6月30日期间在该国南部八家社区卫生诊所连续接受免散瞳数字视网膜摄影的所有6962例2型糖尿病患者。还对接受免散瞳视网膜摄影且同时由眼科医生散瞳检查的患者进行了随机抽样比较。所有患者的平均年龄为64.2岁。共有5960例（占所有照片的85.6%）照片质量足以用于诊断。在1092例（18.3%）患者中发现了任何程度的DR。49.4%的患者眼底照片正常。在32.2%的病例中发现了非DR病变。在发现DR的病例中，73.3%（801例）患有轻度非增殖性视网膜病变（NPDR），7.1%（77例）患有中度NPDR，6.8%（74例）患有增殖性视网膜病变，12.8%（140例）患有糖尿病性黄斑水肿。老年患者拍摄质量差照片的可能性更大（70岁及以上患者与70岁以下患者之间p<0.001）。当将免散瞳眼底摄影与眼科医生的散瞳眼底检查进行比较时，发现敏感性为99.3%，特异性为88.3%，阳性预测值为85.3%。免散瞳数字视网膜摄影是一种有效的DR筛查方法。该检测具有高敏感性和特异性。在该国南部社区卫生中心进行的这项检测促成了1000多例DR的早期诊断。许多患者可以快速有效地得到随访，尽管该检测主要在老年人中不能替代散瞳后的全面眼部检查。其他威胁视力的非DR眼部病变也可以被检测到。
