Predicting Development of Proliferative Diabetic Retinopathy

OBJECTIVE

Identifying individuals most at risk for diabetic retinopathy progression and intervening early can limit vision loss and reduce the costs associated with managing more advanced disease. The purpose of this study was to identify factors associated with progression from nonproliferative diabetic retinopathy (NPDR) to proliferative diabetic retinopathy (PDR).

RESEARCH DESIGN AND METHODS

This was a retrospective cohort analysis using a claims database of all eye care recipients age ≥30 years enrolled in a large managed-care network from 2001 to 2009. Individuals with newly diagnosed NPDR were followed longitudinally. Multivariable Cox regression analyses identified factors associated with progression to PDR. Three- and five-year probabilities of retinopathy progression were determined.

RESULTS

Among the 4,617 enrollees with incident NPDR, 307 (6.6%) developed PDR. After adjustment for confounders, every 1-point increase in HbA_1c was associated with a 14% (adjusted hazard ratio 1.14 [95% CI 1.07–1.21]) increased hazard of developing PDR. Those with nonhealing ulcers had a 54% (1.54 [1.15–2.07]) increased hazard of progressing to PDR, and enrollees with nephropathy had a marginally significant increased hazard of progressing to PDR (1.29 [0.99–1.67]) relative to those without these conditions. The 5-year probability of progression for low-risk individuals with NPDR was 5% (range 2–8) and for high-risk patients was 38% (14–55).

CONCLUSIONS

Along with glycemic control, nonophthalmologic manifestations of diabetes mellitus (e.g., nephropathy and nonhealing ulcers) are associated with an increased risk of diabetic retinopathy progression. Our retinopathy progression risk score can help clinicians stratify patients who are most at risk for disease progression.Diabetic retinopathy is the leading cause of new cases of legal blindness in the U.S. (1), affecting 4.2 million Americans, 655,000 of whom have sight-threatening retinopathy (1,2). Identifying patients who are at increased risk of progression from nonproliferative (NPDR) to proliferative diabetic retinopathy (PDR) is important for many reasons. From the patient’s perspective, individuals who progress from NPDR to PDR frequently experience a decline in best-corrected visual acuity, which can have a profound impact on health-related quality of life (3). In addition, those who develop PDR are at substantially increased risk of serious complications that can result in permanent vision loss such as tractional retinal detachment, vitreous hemorrhage, and neovascular glaucoma (4,5). From a societal perspective, the costs of caring for patients with PDR are four times greater than the costs of managing patients with NPDR. One study found the average cost of caring for patients with NPDR to be 292 USD, while it cost 1,207 USD to manage patients who develop PDR (6). Another study conducted by the National Health Services in Taiwan found that individuals who progressed from NPDR to PDR were noted to have an increase in expenditures of 3,482 USD (7). The ability for clinicians to identify and treat patients early in the disease process, before they experience progression to PDR, may result in considerable cost savings, especially in light of the growing number of individuals with diabetes mellitus (DM) in the U.S. population.In patients with DM, metabolic control as measured by HbA_1c and disease duration account for only 11% of the risk of retinopathy, leaving 89% to other factors (8). Several large population-based studies including the Wisconsin Epidemiologic Study of Diabetic Retinopathy (WESDR), the UK Prospective Diabetes Study (UKPDS), and the Action to Control Cardiovascular Risk in Diabetes (ACCORD) study have identified other risk factors associated with the development or progression of diabetic retinopathy (9–11). From the results of these studies, age, sex, socioeconomic status, and comorbid systemic arterial hypertension are considered important determinants of retinopathy risk. We are unaware of any studies in the literature that have integrated these and other factors into a comprehensive diabetic retinopathy risk score that can help clinicians identify individuals who are at increased risk of progression from NPDR to PDR. Risk calculators such as the Framingham Risk Score for Atrial Fibrillation (12) and the Ocular Hypertension Treatment Study risk calculator (13) have been found to be useful in aiding clinicians with patient decision making.The purpose of this analysis was to assess risk factors associated with progression of diabetic retinopathy among a diverse group of individuals with DM enrolled in a large managed-care network. By following beneficiaries longitudinally, we sought to confirm previously identified risk factors and to define additional risk factors that may be associated with progression from NPDR to PDR. Finally, using the risk factors identified from our regression models, we developed a risk score that clinicians can use to identify groups of individuals who are at low and high risk of retinopathy progression.