Implementation of a pragmatic pathway to ensure high-quality processing of heterogeneous remote health sensing data

Remote health monitoring systems (RHMS) can improve access to care and health outcomes by enabling continuous monitoring of an individual's overall health status outside traditional clinical settings. However, ensuring data quality and effectively cleaning and processing complex, heterogeneous RHMS sensing data collected in naturalistic environments can present unique challenges. This article presents a pragmatic approach to clean and process RHMS data received from multiple devices/sources (e.g., wearables, ambient environmental sensors) in various formats (e. g., JSON, CSV) during real-word deployments. Our proposed pathway is highly relevant given the growing use of RHMS to support home-based healthcare and provides a useful guide for others engaged in similar research. The pathway describes a series of key steps implemented by our interdisciplinary team, including data preparation, processing, enrichment, and quality assurance, all designed to ensure that the highest quality data are available for analysis. Each step is illustrated through examples from deployments of the Behavioral and Environmental Sensing and Intervention for Cancer (BESI-C), a remote health monitoring system designed to empower patients with advanced cancer and their caregivers to monitor cancer pain and distress at home. This paper fills an important gap in the literature by focusing on the "critical middle" of the RHMS data life cycle (data extraction to preparation for analysis) to ensure accurate conclusions are drawn from data output. Our proposed pathway is applicable to other RHMS, across various patient populations and contexts, that collect diverse data streams in real-world settings and offers a strategy to address those challenges and optimize health outcomes.