Toward Personalized Decision Support Systems for Type 1 Diabetes: Integrating Psychobehavioral Factors and Glycemic Control.

Abstract

Background: While novel technologies have improved glycemic control in type 1 diabetes (T1D), their design is often glucose-centric and overlooks critical psychobehavioral elements such as individual treatment preferences and therapeutic goals. This article presents an algorithm development framework and study layout aimed at incorporating psychobehavioral factors into the design of technology for the management of T1D.

Methods/design: A decision support system (DSS) was engineered at the University of Virginia providing therapeutic advice to individuals with T1D regarding optimal insulin dosing parameters, bolus calculation, safe undertaking of physical activity, and risk for hypoglycemia during the day as well as at bedtime. To accommodate individual preferences as to how the therapeutic advice is delivered, the DSS was designed with two possible operating modalities: prescriptive-offering optimized therapeutic recommendations for structured guidance, and informative-providing users with actionable insights to support informed decision-making. To test the DSS, a randomized crossover clinical trial design is described, where participants are sequentially exposed to each modality in random order. Throughout the study, glycemic control is captured by continuous glucose monitors, while patient-reported outcomes are assessed through psychometric evaluations.

Conclusion: This work introduces a novel framework for the design of personalized DSS technology capable of tailoring the delivery mode of therapeutic insights to each user's preferences. By integrating psychobehavioral factors into algorithm design, this work seeks to advance the development of adaptive, user-centric technologies that can enhance both clinical outcomes and patient quality of life.