Moderator of 1 Session
Presenter of 4 Presentations
Using Digital Health Technology to Prevent and Treat Diabetes
The Use of AI, Machine Learning, and Digital Therapeutics for Behavioral Modification
Big data in diabetes centers and hospitals – what do we do with it?
Abstract
Abstract Body
Diabetes centers, healthcare systems, and individuals with diabetes generate many types of data about diabetes-related outcomes and about self-management behaviors, comorbid medical conditions, and clinical care-related events. Yet only a small fraction of these data are used by clinicians regularly for decisionmaking. Risk-based management protocols can help diabetes centers to improve both the quality and cost-efficiency of care. These protocols may be driven by biomarkers of risk extracted or derived from electronic health records, diabetes self-management devices (or the cloud services that receive their data), and digital patient reported outcomes platforms; protocols may alternately be driven by forecasting of negative outcomes via Artificial Intelligence/Machine Learning approaches. The participation by diabetes centers in Learning Health Networks, with data sharing to a central data repository, can accelerate Big-Data-driven quality-improvement of care delivery. The presenter will review examples of risk-based management approaches using each technique, including novel biomarker-based risk indices like the 6 Habits of self-management, the Diabetes Care Index, and the Risk Indexes for Poor Glycemic Control and for Diabetic Ketoacidosis (RI-PGC and RI-DKA, respectively). The presenter will further examine the current state of algorithms and AI/ML to manage population health in diabetes clinics, including a population health dashboard to reduce deterioration in glycemic control in the post-diagnostic period for type 1 diabetes, a precision medicine project for type 2 diabetes incorporating multiple -omics biomarkers, and the Rising T1DE Alliance, which seeks to implement multiple ML models to predict outcomes in clinical care, and to test remote patient monitoring along with multiple digital and behavorial health interventions to improve those predicted outcomes via a risk-based management approach.
REMOTE PATIENT MONITORING IN YOUTH WITH TYPE 1 DIABETES (T1D) PREDICTED TO EXPERIENCE A RISE IN A1C%: COMPARISON TO A CLINIC-DERIVED, PROPENSITY SCORE-MATCHED CONTROLS
Abstract
Background and Aims
To determine if 1) remote patient monitoring (RPM) can improve glycemic control among youth with T1D who are predicted to experience a rise in A1c% using logistic regression and 2) propensity-score (PS) matching (PSM) RPM youth to non-RPM youth can control for confounding bias.
Methods
We collected data from 2-18-year-olds attending a tertiary care network of diabetes clinics in the Midwestern USA from 11/2018-9/2021. Eligible youth had baseline A1c% ≥7.2, predicted 90-day rise in A1c% ≥0.3 via advanced machine learning, and follow-up A1c% measured 70-180 days after baseline. Criteria used to calculate PS included sex, ethnicity, race, insurance, technologies, age, T1D duration, baseline A1c%, and predicted 90-day A1c% change. We compared each RPM-youth with three matched, non-RPM youth identified using a PS within ±0.05.
Results
We matched 201 non-RPM youth to 67 RPM youth. The final cohort was 60% female, 4% Hispanic, 76% White, 54% private insurance, 37% on CGM and insulin pump, median age 13.4 years (IQR=10.3,16.0), T1D duration 44.6 months (17.6,84.7), baseline A1c% 7.9 (7.5,8.8), and predicted 90-day A1c% change 0.39 (0.33,0.48). After PSM, we found no significant differences for RPM and non-RPM youth (p’s=0.10-0.99), suggesting this method may be appropriate for creating a balanced clinic-derived control sample. However, 64% of RPM youth experienced no rise in A1c ≥0.3% compared to 53% of non-RPM youth (p=0.10).
Conclusions
Youth receiving RPM may experience improved glycemic control relative to a clinic-derived control sample based on PSM, but these results require verification in a large multisite traditionally controlled study.