iMatter2: An AI-driven approach to supercharge a novel digital patient-reported outcomes tool for diabetes management

点击翻译按钮获取中文摘要

Project Summary: Uncontrolled type 2 diabetes (T2D) is a major health problem in the US that constitutes a significant cause of morbidity and mortality, particularly in minoritized populations who continue to suffer disproportionately higher rates of complications. Despite the significant role primary care practices play in the management of T2D, they were never designed nor ideally suited to capture and address the complex psychosocial (e.g., stress) and behavioral factors (e.g., adherence to self-care) that significantly impact glycemic control. However, without incorporating the psychosocial and behavioral impacts of T2D into clinical decision-making, achievement of the outcomes desired by patients and primary care providers (PCP) will be unattainable. Application of patient-reported outcomes (PROs) to T2D care represents an ideal opportunity to capture the psychosocial and behavioral impacts of living with T2D on patients' clinical outcomes. In direct response to NOT-HS-16-015, this renewal application, which builds on our AHRQ-funded randomized control trial (RCT) evaluating the efficacy of an innovative mHealth PRO tool, will address this gap in the literature. Using a theory-driven and user-centered design approach, we will conduct the study in two phases: 1) a formative phase to refine and user-test iMatter2 in real-world settings; and 2) a clinical trial phase. The formative phase will use a mixed-methods approach to: a) integrate enhanced functionalities into our existing tool and the electronic health record (EHR) in a diverse network of primary care practices serving minoritized populations; and b) evaluate the usability of iMatter2 in a subset of patients with T2D and PCPs to optimize the tool's performance and workflow integration. For the clinical trial phase, we will evaluate in a hybrid type 1 RCT, the effectiveness of iMatter2 vs usual care (UC) on HbA1c reduction at 12-months among 353 patients with uncontrolled T2D (HbA1c>7%). Using the extended RE-AIM framework, we will apply an equity lens to measure the reach, adoption, and implementation of iMatter2. We will also explore the associations between patient's PRO responses and HbA1c reduction. PCPs will be the unit of randomization with all patients within the PCP's panel in the same group. Patients randomized to iMatter2 will receive and respond to personalized PROs via text message, receive personalized motivational and educational messages via an AI chatbot; and have access to an interactive dashboard that visualizes their daily PRO and HbA1c data. PCPs randomized to iMatter2 will have access to EHR-integrated clinical decision support tools to nudge PCPs to view the PRO reports and HbA1c data. Patients randomized to UC will receive standard T2D treatment recommendations from their PCPs; PCPs will not have access to the EHR reports. Primary outcome is mean reduction in HbA1c, extracted from home A1c kits, from baseline to 12 months. Secondary outcomes are 1) equitable reach of iMatter2 across settings and patients; 2) equitable adoption of iMatter2 across PCPs and patients; and 3) equitable engagement (implementation) with iMatter2 across PCPs and patients.