Project Details -
Completed
-
Grant NumberR01 HS025429
-
Funding Mechanism(s)
-
AHRQ Funded Amount$1,957,024
-
Principal Investigator(s)
-
Organization
-
LocationBerkeleyCalifornia
-
Project Dates09/01/2017 - 06/30/2022
-
Care Setting
-
Medical Condition
-
Type of Care
-
Health Care Theme
Diabetes and depression are major public health problems that disproportionately affect low-income individuals and racial and ethnic minorities. While these conditions often occur together and should be addressed concurrently, interventions are primarily delivered by separate clinicians with separate, condition-specific treatments that rely heavily on clinician contact. Mobile health (mHealth) interventions can improve self-management of both using behavioral activation approaches, and are cost effective and feasible methods for delivering support. Existing mHealth interventions have shown preliminary success but have had difficulty sustaining engagement. Patients have been found to be most engaged when they perceive mobile interventions are coming from trusted clinicians, and are more likely to send self-management data to clinicians when they believe it will be used to improve their care.
The Diabetes and Mental Health Adaptive Notification Tracking and Evaluation (DIAMANTE) study will implement and evaluate an adaptive-learning, clinic-integrated, mobile intervention targeting physical activity to manage diabetes and depression in low-income minority patients. This project will use machine learning algorithms to adapt and deliver health messages via text messaging to motivate these individuals based on their needs and motivations.
The specific aims of the project are as follows:
- Conduct User-Centered Design of the DIAMANTE intervention in English and Spanish.
- Evaluate the effectiveness of adaptive versus static messaging.
- Evaluate the effectiveness of nurse phone outreach for non-responsive participants.
Researchers will evaluate a text message intervention with an adaptive machine learning algorithm that learns from patient step count data and patient-entered blood glucose and mood ratings; data will be compared to a control group receiving static text messages. Investigators will then re-randomize non-responsive participants from both arms to receive one-on-one nurse outreach using a Sequential, Multiple Assignment, Randomized Trial (SMART) design. Leveraging the SMART design will allow them to test the effectiveness of the costlier clinical outreach among a subsample who have been identified as needing additional support.
Data from the study will help determine if text messaging can be implemented to help improve treatment of chronic illness in a primary care setting. The results will also provide information on the impact of personalizing content using machine learning algorithms, as well as the impact of providing clinician support for those receiving mobile health interventions. Findings can be applied to future interventions seeking to employ text messaging to improve adherence to treatments for chronic disorders and will aid in the development and dissemination of mobile health tools for chronic illness management in diverse patients.