Personalized Disease Screening Decisions Considering a Chronic Condition

Clinical practice guidelines do not sufficiently address the needs of patients with chronic conditions as these guidelines focus on single disease management and ignore unique patient-specific conditions. As a result, a nonpersonalized approach to the management of the patients with chronic conditions leads to adverse events and increases the financial burden on the healthcare system as over 150 million Americans experience chronic conditions. To this end, we develop a stochastic modeling framework to personalize the disease screening decisions for patients with or at risk for developing a chronic condition and provide an exact solution algorithm. We consider the optimal management of the screening decisions for an index disease (e.g., breast cancer, colorectal cancer, human immunodeficiency virus, etc.) while accounting for the existence of a chronic condition (e.g., hypertension, diabetes, Alzheimer's disease, etc.). Our modeling framework is particularly useful for the cases where the chronic condition affects the risk of the index disease. In a case study using real breast cancer epidemiology data, we demonstrate how our modeling framework can be used to personalize breast cancer screening for women with type 2 diabetes. In addition to providing a personalized breast cancer screening schedule for women with diabetes, we find some important policy insights that were not previously recognized by the medical community. More specifically, we find that compared with women without diabetes, women with diabetes should be screened less aggressively, but screening should end at similar ages. We also find that adherence to the optimal screening policy is more crucial for women with diabetes compared with nondiabetic women. Our main insight on screening recommendations also has important resource implications as it leads to fewer screening mammograms. That is, compared with the current national breast cancer screening guidelines, the optimal breast cancer screening policy for women with diabetes could save the healthcare system approximately 2.6 million mammograms annually.

Automated summary

Current clinical practice guidelines lack the ability to address the personalized needs of patients with chronic conditions, resulting in adverse events and increased financial burden on the healthcare system. In order to solve this problem, a stochastic modeling framework is developed to personalize disease screening decisions for patients with or at risk of developing chronic conditions. This framework proves useful when the chronic condition affects the risk of the index disease.