Export Data


Background Diabetes is a major cause of avoidable blindness in both developing and developed countries. The prevalence of retinopathy in diabetics increases with the duration of the disease and patients with Diabetes are 25 times more likely to become blind than non-diabetics. Therefore, there is definitely a need for an effective screening strategy to identify early changes of diabetic retinopathy (DR) before it progresses to a sight threatening stage. The smartphone which has pervaded our medical practice in recent years is portable and can be used in diabetic screening clinics to document and monitor diabetic retinopathy changes. Objective To determine the sensitivity and specificity of a smartphone (Samsung galaxy S6 Edge) in diagnosing and staging diabetic retinopathy as well as to describe the clinical presentation of diabetic retinopathy in diabetics attending the Mbarara Regional Referral Hospital. Methods This was an analytical, hospital based cross sectional study. There were 207 patients enrolled in the study and after pupillary dilation, all patients underwent examination with the indirect ophthalmoscope and then had 7 field smartphone funduscopy done at the General Medical clinic in MRRH, Mbarara, Uganda. The diagnosis and grading of smartphone photos were done by two Independent retinal specialists who used the International Clinical Diabetic Retinopathy Disease severity scale. Sensitivity and specificity of smartphone photographs was analysed using indirect ophthalmoscopy as the gold standard. xix Results Sensitivity of smartphone in diagnosing DR was observed to be 70% and specificity 94%. The sensitivity and specificity of the smartphone in grading Proliferative Diabetic Retinopathy was 100% and 99.4%. For macular edema, sensitivity was 38.5% and specificity 97.9%. Kappa agreement was 0.86 in diagnosing DR and 0.84 in grading diabetic retinopathy. The prevalence of DR was 13.5% and the proportion of DR stages were: 11.6% with NPDR, 1.9% with PDR and 6.8% with macular edema. The proportion of visual impairment was 9.6% and blindness 0.5% and four patients already had complications of DR at the time of enrollment: retinal detachment, retinal vein occlusion, subhyaloid hemorrhage and an only eye with PDR and vitreomacular traction. Conclusion Due to the low sensitivity of smartphone photography in diagnosing DR, further studies should be done with better skilled personnel to make conclusive statements about this innovation. Also, there is need to institute an active DR screening clinic at MRRH, which is equipped with a fundus camera and with trained Ophthamic personnel to do early screening of all diabetics.

Charts showing individual downloads annually

Line Graph showing individual downloads annually