A0741
Title: Dealing with zero-inflated count data in mobile health
Authors: Xueqing Liu - Duke-NUS Medical School (Singapore) [presenting]
Nina Deliu - Sapienza University of Rome; University of Cambridge (Italy)
Bell Lauren - University of Cambridge (United Kingdom)
Bibhas Chakraborty - Duke-NUS Medical School, National University of Singapore (Singapore)
Abstract: Mobile health technologies play a crucial role in enhancing distal outcomes, such as clinical conditions, by optimizing proximal outcomes through just-in-time adaptive interventions. Micro-randomized trials are considered the gold standard for constructing these interventions. In this design, participants are sequentially randomized to different intervention options. This work is motivated by the Drink Less micro-randomized trial, where the proximal outcome is a longitudinal count outcome. Analyzing data from micro-randomized trials often revolves around two critical scientific questions:(1) Which intervention can influence the proximal outcome? (2) In which context should the intervention be delivered? To address these questions, we introduce the concept of causal excursion effects and propose novel methods for estimating such effects. Through comprehensive simulation studies, we evaluate the performance of the proposed estimators. Additionally, we apply our method to real-world data from the Drink Less trial, highlighting its practical utility in the mHealth domain.
