ABSTRACT

Background:

Type 2 diabetes mellitus (T2DM) significantly impacts public health, with approximately 21 million US adults diagnosed. Telehealth interventions show promise for improving T2DM outcomes through digital self-management techniques, but face challenges due to disparities in digital literacy and access, especially in rural areas. There is a need for sustainable T2DM management interventions that require minimal digital literacy and are widely accessible. We propose an innovative, individualized lifestyle modification intervention delivered via standard phone service to control blood glucose levels in individuals with T2DM.

Objective:

This paper outlines the protocol of a pilot study aiming to assess the feasibility of implementing and preliminary effectiveness of an artificial intelligence-assisted individualized lifestyle modification intervention for glycemic control in rural populations delivered via landline telephone service.

Methods:

This study employs a multiphase optimization strategy (MOST) and includes two experimental intervention components: automated vs. human health coaching and adapted vs. fixed gamified reward levels based on daily automated monitoring calls. We aim to recruit 88 patients with diabetes and HbA1C levels 6.5–11.5%. Participants receive daily behavioral monitoring phone calls to evaluate self-management behaviors. Participants also receive either weekly human health coaching or automated AI-driven health coaching for six months. In the fixed-reward arm, participants earn 60 cents per day for answering daily calls, while in the adapted gamified reward arm, rewards start at 20 cents per day and increase weekly, with penalties for missed days. Both arms can earn up to $100.80 over six months. Semi-structured exit interviews will gather patient insights post-trial. Primary outcomes include feasibility measures, HbA1c levels, and lipid profiles.

Results:

We have screened 813 people with diabetes and enrolled 54 participants since the launch of the study. We project that enrollment and analyses to assess feasibility completed in 2025.

Conclusions:

This intervention lays the groundwork for a future optimization trial addressing T2DM management, reaching populations through digital health while requiring minimal digital skills. It has the potential to be a scalable low-cost AI-assisted diabetes management solution that is accessible to rural communities and those with low digital literacy or smartphone access. Clinical Trial: ClinicalTrials.gov Identifier: NCT05344859