作者: David J. Couper , Ronald Klein , Larry D. Hubbard , Tien Yin Wong , Paul D. Sorlie
DOI: 10.1016/S0002-9394(01)01315-0
关键词:
摘要: Abstract PURPOSE: Retinal microvascular characteristics, as graded from retinal photography, have been shown to predict stroke. We evaluated the reliability of photographic grading in Atherosclerosis Risk Communities Study. DESIGN: Cohort study. METHODS: photographs were taken all subjects who attended third Study examination (1993 1995). These using standardized protocols. Focal characteristics a “light box” system. Generalized arteriolar narrowing was quantified computer-assisted measurements digitized photographs. Two substudies conducted investigate these methods. In Individual Variability Study, selected (n = 206) had two on one day, and further or 3 weeks later. Grader stratified random sample repeat 495 for light box grading; n 276 grading). RESULTS: Reliability quantification generalized high both studies (reliability coefficients 0.64 0.69 0.79 0.83 Study). There more variability focal abnormalities (same individuals, photographs) tended be greater than photographs, gradings). CONCLUSION: especially narrowing, can ascertained reliably by methods, supporting validity their associations with cardiovascular disease. However, appear vary somewhat between eyes over time single individual.