Reduced vessel density of the choriocapillaris was one independent risk factor for visual impairment in a DR nomogram. The others were diabetic macular ischemia grade, disorganization of the retinal inner layers and outer layer disruption. Photo: Julie Rodman, OD, MSc. Click image to enlarge.
Retinal structural abnormalities are associated with diabetic retinopathy (DR) severity and vision loss. In a recently published paper, researchers scored various retinal risk factors correlated with vision loss in DR to create a nomogram that predicts visual impairment.
They retrospectively enrolled patients who had undergone OCT-A, fundus fluorescein angiography and swept-source OCT of the macula to observe retinal blood flow, stage disease and define retinal structures. The researchers considered visual acuity ≥0.5 logMAR as impaired. They screened characteristics that correlated with visual acuity using binary logistic regression. These factors were input into a multivariate binary stepwise regression to create a nomogram, which was then validated.
In total, the researchers analyzed 29 parameters and 13 characteristics for their model. They reported that the following were statistically significant: diabetic macular ischemia grade, disorganization of the retinal inner layers, outer layer disruption and vessel density of the inferior aspect of the choriocapillaris layer.
The validated model had an area under the curve of 0.931. The researchers noted in their paper that analysis confirmed that risk threshold probabilities can be used as clinical practice guides. The clinical impact curve of the model shows the proportion of individuals at risk at each threshold probability.
Importantly, the study used patient data from a single visit, so the “model can only predict the current risk of visual impairment and not changes in visual acuity at a future time,” the researchers pointed out in their paper. They recommended future prospective studies with external validation to improve the model.
Zhao Y, Yu R, Sun C, et al. Nomogram model predicts the risk of visual impairment in diabetic retinopathy: a retrospective study. BMC Ophthalmol. 2022;22:478.