Session Title: Quick Fire Free Paper Session 04
Session Date/Time: Sunday 29/09/2013 | 11:00-13:00
Paper Time: 12:00
Venue: Hall C (Level 1)
First Author: A.Ranganath UK
Co Author(s): J. Sardar P. Moulik A. Macleod
The purpose of our study was to examine prognosis and prognostic indicators in diabetic patients with retinal emboli compared to diabetic patients of the same age who were screened on the same day but with no retinal emboli. We analysed this age-matched group of people with diabetes for different variables including mortality, grading of diabetic retinopathy, blood HbA1C, creatinine and lipid levels. We aimed to find out if patients with retinal emboli had associated risk factors that need to be addressed.
Data was gathered from diabetic patients attending the Shropshire Diabetic Eye Screening Programme (DESP) from 2008 to 2012. Patients are screened annually, both in the community and in the two District General Hospitals.
Retrospective case-control study of all 133 diabetic patients who attended the annual Diabetic Eye Screening Programme from 2008 to 2012, and were identified as having retinal emboli from fundus images, against 133 diabetic control patients without evidence of emboli, screened on the same day with the nearest age to the cases. The data was collected from the eye screening database of all diabetic patients attending the Shropshire DESP graded over the time period from 2008 to 2012. Mortality was assessed as of December 2012. Trained graders from the English National Screening Programme for Diabetic Retinopathy (ENSPDR) graded the fundus images according to the latest National Screening Committee criteria. They were instructed to look for retinal emboli and to refer all patients identified with retinal emboli to nominated physicians for further management. The fundus images were further studied by a Consultant Ophthalmologist to identify the type of emboli(cholesterol/platelet/calcium). Statistical analysis: discrete variables were assessed for statistical significance using the Fishers exact 2-tailed test, and continuous variables using the Paired t-test.
• 17,600 and patients were screened under DESP in Shropshire per year and those with retinal emboli comprised 0.75%. • Type of emboli: Cholesterol – 103 (77.5%), Platelet – 10 (7.5%) and Calcific – 20 (15%) • Location of emboli: Periphery (2nd bifurcation and beyond) – 67 (50.3%) and Central – 66 (49.6%) • The average age in years in the cases group was 72.50 and in the control group 72.71 with p-value = 0.88 (paired t-test). • Mortality among cases was 27 and controls 13, giving a statistically significant p-value = 0.0249 (fisher exact 2 tailed test). • The average duration of diabetes mellitus was 10.69 years among cases and 10.52 years among the controls, with p-value = 0.85 • Average HbA1c (mmol/mol) in cases was 57.83 and controls 55.91 with p-value = 0.33 • Average creatinine (Umol/L) among cases was 108.59 and controls 87.23 with a statistically significant p-value = 0.00 • Average HDL cholesterol (mmol/L) among cases is 1.24 & controls 1.32 with p-value = 0.07. • Lastly, the average LDL cholesterol (mmol/L) among cases is 1.92 and controls 2.0 with p-value = 0.38
Despite what we believe to be careful analysis, those with retinal emboli had twice the mortality thus reaching statistical significance. Renal function was significantly worse in patients with retinal emboli, presumably due to reno-vascular disease. Lower levels of HDL cholesterol would be expected in a high-risk cardiovascular group. Absence of LDL cholesterol relation might be explained by treatment with cholesterol lowering agents especially statins. We therefore strongly believe that the presence of retinal emboli cannot be ignored in an eye-screening programme. In summary, we would like to stress focussing on cardiovascular risk prevention in this very high-risk group of diabetic patients."