An Adaptive Optimization Approach to Personalized Financial Incentives In Behavioral Interventions

Obesity is a critical healthcare issue affecting the United States. The least risk treatments available for obesity are behavioral interventions meant to promote diet and exercise. Often these interventions contain a mobile component that allows interventionists to collect participant-level data and provide participants with incentives and goals to promote long-term behavioral change. Recently, there has been interest in using direct financial incentives to promote behavior change. In general these are operationalized by having a single intervention budget distributed equally across all participants, meaning intervention costs scale linearly with enrollment. However, as each participant will react differently to different incentive structures and amounts, there has been recent interest by clinicians in using personalized models to more effectively allocate intervention budgets. The key challenge for personalization is that the clinicians do not know a priori which participants will be most responsive to these incentives, and moreover how to administer their scarce budget to maximize cohort health outcomes. In this paper, we consider this challenge of designing personalized weight loss interventions that use direct financial incentives to motivate weight loss while remaining within a budget. We create a predictive model that can predict how individuals may react to different incentive schedules within the context of a behavioral intervention. We integrated this predictive model in an adaptive optimization framework that over the course of the intervention computes what incentives to disburse to participants and remain within the study budget. We show that our optimization framework is asymptotically optimal. We demonstrate the effectiveness of our approach using real-world data from a real world weight loss trial that used financial incentives to incentives weight loss. Our results show that using an 60–80% smaller budget, our adaptive optimization framework is able to help the same number of participants lose weight as an existing one-size-fits-all intervention. Furthermore, using our adaptive optimization framework would spend only $6 per individual across 26 weeks as opposed to the current intervention which would spend $42 per individual in the same time frame to achieve similar clinical outcomes. Our results highlight that providers who choose to implement behavioral interventions at scale will need to use a personalized approach to effectively use their limited budgets.

• 저널Production and Operations Management · 35(2) · 546–565
• 토픽Mental Health Research Topics · Experimental and Cognitive Psychology
• DOI10.1177/10591478251349391
• 저자Qiaomei Li