Combination Therapy Devices in Diabetes Management

Abstract

Diabetes management has evolved from isolated insulin injections and intermittent glucose checks to integrated systems combining continuous glucose monitoring, automated insulin delivery, and digital therapeutic support within unified platforms. Combination therapy devices -- hybrid closed-loop insulin pumps, dual-hormone artificial pancreas systems, sensor-augmented pen injectors, and AI-driven multi-therapy optimisers -- represent a paradigm shift toward holistic glycaemic management that coordinates monitoring, drug delivery, and behavioural intervention. This study presents the Combination Therapy Device Assessment Framework for Diabetes (CTDAF), evaluating five device integration levels - standalone CGM with manual insulin adjustment, sensor-augmented pump therapy with low-glucose suspend, hybrid closed-loop single-hormone systems, dual-hormone closed-loop systems (insulin + glucagon), and AI-integrated multi-therapy platforms combining insulin delivery with GLP-1 dosing guidance and behavioural coaching -- across four glycaemic management tasks: time-in-range optimisation, hypoglycaemia prevention, glycaemic variability reduction, and patient burden minimisation. Our Diabetes Device Integration Score (DDIS) integrates time-in-range achievement, hypoglycaemia avoidance, variability control, device burden, and treatment satisfaction. AI-integrated multi-therapy platforms achieved the highest DDIS (0.928) through reinforcement learning algorithms that simultaneously optimised basal insulin, bolus recommendations, and GLP-1 timing to achieve 82% time-in-range with 1.2% time-below-range, while dual-hormone systems achieved the highest hypoglycaemia prevention (0.960) through automated mini-dose glucagon delivery that reduced time below 54 mg/dL by 78%.