Abstract
Purpose :
Diabetic retinopathy (DR) regression is infrequently reported, but a few studies suggest DR regression occurs in patients with mild NPDR (ETDRS Level 20), hemoglobin A1c of 6-7% (A1c), and shorter duration of diabetes (10 yrs). Our study estimates how many patients demonstrated improvement in mild NPDR (ETDRS Level 35) using a nationally-distributed cohort examined by the Indian Health Service (IHS) teleophthalmology program from 2015-2020.
Methods :
This is a retrospective study of deidentified electronic medical records. Patients underwent ultrawide field imaging at primary care sites. Eligible patients had a baseline evaluation in 2015-2017 with a diagnosis of mild NPDR and at least one follow-up (FU) visit within 3 years of their baseline exam (N=2,910). The analyses calculated the numbers and percentages of patients with varying degrees of NPDR or proliferative DR at their final FU examination, parsed by timing of their final FU visit. Ordinal logistic regression models estimated the net effects of age, gender, months from baseline exam, diabetes (DM) duration, DM therapy, baseline A1c and percent change in A1c over time on FU DR diagnosis.
Results :
In patients seen 12-29 months after baseline, the most common final FU diagnosis was no DR (~41% of patients within each group defined by timing of FU; table below), indicating improvement. Without taking timing of the final exam into consideration, 40% of patients had no DR at their last FU visit within the study’s timeframe. A unit increase in years of age (OR=1.01, confidence interval [CI]=1.00-1.02), months from baseline exam (OR=0.99, CI=0.98-1.00), and percent change in A1c (OR=1.00, CI=0.99-1.00) slightly increased the odds of DR regression at FU. Longer duration of DM (e.g., 16+ vs. 1-5 years [OR=0.42, CI=0.42-0.53]), an insulin prescription (vs. diet only [OR=0.54, CI=0.36-0.83]) and higher baseline A1c (OR=0.83, CI=0.79-0.86) decreased the odds of DR regression.
Conclusions :
In patients with baseline mild NPDR, there were nearly twice as many regressors as progressors despite higher A1cs and older ages. Thus, healthcare may be optimized by focusing on reducing A1c in younger patients, those taking insulin, and those with longer duration of DM, through allocation of more eye health resources in primary care settings.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.