Predictive factors for patients receiving intravitreal anti-vascular endothelial growth factor for the treatment of diabetic macular edema

Abstract

Purpose:

To determine whether anemia and other demographic or laboratory “risk factors” impact anti-vascular endothelial growth factor treatment in diabetic macular edema patients.

Methods:

This is a retrospective, time-varying cohort study using a medical claims database to identify new diabetic macular edema patients who had received at least one intravitreal injection of anti-vascular endothelial growth factor. Exclusion occurred for having <2 years in the plan prior to diabetic macular edema diagnosis, any history of proliferative retinopathy or any treatment that is used for diabetic macular edema. Covariates of interest were demographic characteristics, laboratory values, and clinical factors such as previous anti-vascular endothelial growth factor used, number of involved eyes, year of treatment, and time since last injection. Those variables that changed with time were assessed and updated at each visit. The main outcome measure was the odds of receiving treatment at any visit.

Results:

In total, 189 new diabetic macular edema patients with follow-up were analyzed, covering 729 visits with 543 (74.5%) receiving treatment. Univariate analysis showed that male gender (odds ratio: 0.54, 95% confidence interval: 0.32–0.91, p = 0.03), every week since last injection (odds ratio: 0.94, 95% confidence interval: 0.91–0.97, p = 0.001), and having two eyes affected (odds ratio: 2.09, 95% confidence interval: 1.10–3.97, p = 0.02) were associated with getting an injection. After multivariate analysis, only time since previous injection with every week that passed reduced the odds on having an injection at the next visit (odds ratio: 0.95, 95% confidence interval: 0.92–0.97, p < 0.001). Anemia was not associated with receiving an injection (odds ratio: 1.05, 95% confidence interval: 0.61–1.80, p = 0.86).

Conclusion:

This study used time-varying methodology to better identify which patients will likely need an injection at any one visit. While anemia was not found to impact injections, our results can aid future endeavors that may incorporate clinical visit information in developing a full prediction model to help make diabetic macular edema care more efficient.