Screening for Presence or Absence of Diabetic Retinopathy: A Meta-analysis

摘要: OBJECTIVES: To examine how mydriasis and the medical qualifications of photographers who take retinal photographs influence accuracy screening for diabetic retinopathy (DR). METHODS: Our meta-analysis included studies that measured sensitivity specificity tests designed to detect any DR, sight-threatening or macular edema. Using random-effects logistic regression, we examined effect variations in mydriatic status on specificity. RESULTS: Only category "any DR" had sufficient consistency definition, number (n = 20), assessments (n = 40) meta-analysis. Variations did not significantly (odds ratio [OR], 0.89; 95% confidence interval [CI], 0.56-1.41; P = .61) (OR, 0.94; CI, 0.57-1.54; P = .80). 1.25; 0.31-5.12; P = .75). Specificity detection DR was higher methods use a photographer with specialist eye 3.86; 1.78-8.37; P = .001). CONCLUSION: Outreach is an effective alternative on-site examination. It has potential increase coverage high-risk patients remote resource-poor settings without risk missing opportunity prevent vision loss. analysis confined presence absence DR. Future should consistent classification schemes facilitate further analysis.