Evaluating Screening Policies for Diabetic Retinopathy: A Simulation Approach

Diabetic retinopathy (DR) is the leading cause of blindness in American Adults and the most common diabetic eye disease. More than 60% of patients with type 2 diabetes and more than 90% of patients with type 1 diabetes will develop DR within 20 years of diagnosis. Regular and timely screening examinations are crucial as early treatment can prevent up to 98% of DR-related vision loss. Unfortunately, only 30% to 60% of the patients are screened on a yearly basis. The current clinical screening exams are time-consuming, inconvenient, and costly, leading to such low compliance rates. In the last decade, teleretinal screening has received increasing attention as an inexpensive and convenient screening technique. However, there is a lack of quantitative understanding about the right balance between teleretinal and in-clinic exams for patients with different characteristics, and how this knowledge can be used for modeling a personalized teleretinal screening policy and integrating it with the existing systems. The goal of this research is to find an optimal, personalized timing of screening exams for patients. To do so, a discrete event simulation model was developed via a Markov chain that replicates the progression of diabetic retinopathy. From the five different screening policies analyzed, the results showed that annual telescreening emerged as a preferable alternative for the base population due to the significantly lower cost. However, when a more at-risk population is evaluated, the annual in-clinic policy showed to be more beneficial to all other policies options tested.