NudgeRank: Digital Algorithmic Nudging for Personalized Health

TL;DR

Rigorous evaluation reveals statistically significant improvements in health outcomes, including a 6.17% increase in daily steps and 7.61% more exercise minutes, as well as user engagement and program enrollment surged.

Abstract

In this paper we describe NudgeRank, an innovative digital algorithmic nudging system designed to foster positive health behaviors on a population-wide scale. Utilizing a novel combination of Graph Neural Networks augmented with an extensible Knowledge Graph, this Recommender System is operational in production, delivering personalized and context-aware nudges to over 1.1 million care recipients daily. This enterprise deployment marks one of the largest AI-driven health behavior change initiatives, accommodating diverse health conditions and wearable devices. Rigorous evaluation reveals statistically significant improvements in health outcomes, including a 6.17% increase in daily steps and 7.61% more exercise minutes. Moreover, user engagement and program enrollment surged, with a 13.1% open rate compared to baseline systems' 4%. Demonstrating scalability and reliability, NudgeRank operates efficiently on commodity compute resources while maintaining automation and observability standards essential for production systems.

Figures (6)

• Figure 1: NudgeRank™ system that automatically generates personalized health nudges for millions of users daily, integrated with mobile app to deliver of nudges via push notifications and collect contextual and nudge engagement data.
• Figure 2: NudgeRank™ Knowledge Graph.
• Figure 3: NudgeRank™ Knowledge Graph Neural Network.
• Figure 4: Evaluation metrics from daily automated model updates, showing the minimum, maximum, mean and standard deviation of Mean Average Precision (map), Normalized Discounted Cumulative Gain @ 3 (ndcg@k), Precision @ 3 (p@k) and Recall @ 3 (r@k).
• Figure 5: Results of execution time versus volume of candidate user-nudge pairs to be scored.