A Comparison between single field nonmydriatic color fundus photography interpreted by family physicians and dilated indirect ophthalmoscopy by ophthalmologists in diabetic retinopathy screening

Objective: To evaluate single field nonmydriatic color fundus photography interpreted by family physician as an alternative method for diabetic retinopathy screening in Thailand. Study design: A diagnostic test (cross-sectional study) Setting: Ophthalmologic outpatient department, Khon Kaen Regional hospital Research methodology: During August to November 2006, 363 type2 diabetic patients who visited ophthalmologic out patient department and fulfilled the inclusion and exclusion criteria were recruited for study. All participants underwent the 2 diagnostic examinations, the first one was the digital fundus camera. The ophthalmologist performed the second diagnostic test, the indirect ophthalmoscopy after dilatation. Results from indirect ophthalmoscopy performed by the ophthalmologists and the digital images interpreted by family physicians were graded by the Proposed International Clinical Diabetic Retinopathy disease severity scales. The validity of the test were calculated, sensitivity and specificity for each family physician (as a dichotomous data, refer or not refer) The referral cut point is at moderate NPDR or greater. The results from indirect ophthalmoscopy by the ophthalmologists were used as reference standard. Results: The prevalence of diabetic retinopathy in this study is 42.1 %.The sensitivity of fundus photographs interpreted by family physician are 64.66, 68.10, 78.45, 93.97, 94.83 % and the specificity are 87.85, 89.47, 79.35, 31.17 and 40.89 %, respectively. Conclusion: This research revealed only fair sensitivity of the test which are not met requirement of a screening test especially in a serious condition.