Digital Biomarkers for Diabetes
Improve patient outcomes and access through target population identification. Our work for a digital therapeutic company had two goals. First was to identify patients on a digital therapeutic who would most benefit from interventions to improve efficiency of providers. Second was to identify patients most likely to benefit from digital therapeutic to improve access with payers.
Research, design, and develop an ML model to predict A1c change after various days on therapy based on digital biomarkers developed from the digital therapeutic, along with a provider-facing product experience. Integrate data from health trackers, clinical systems, patient app usage, self-reported surveys and other sources. Data app to organize provider’s patient panel for optimal triage. Research, design and develop causal models to inform HEOR and identify target patient populations with large payer who are likely to respond to the digital therapeutic.
Insights enabled Type 2 diabetes digital therapeutic currently in clinical trial. Improved target patient access supported strategic partnership between payer and digital therapeutic company. Development of machine learning models for improving patient outcomes published on biomarker development using machine learning.