JMIR Publications

ABSTRACT

Background:

Background:

Health portraits powered by big data integrate diverse health data into actionable insights, supporting precise risk prediction and personalized management of non-communicable diseases (NCDs). Despite their promise, the adoption and application of health portraits remain fragmented, lacking a standardized framework to harness their potential fully.

Objective:

Objective:

This scoping review aims to categorize existing health portrait research in non-communicable diseases management, evaluate the use of big data based on the 3V framework, external validation, and comprehensiveness, and identify challenges, opportunities, and future research directions in this field.

Methods:

Methods:

This study conducted a scoping review following the PRISMA-ScR guidelines and Levac et al.'s 6-step framework. A comprehensive search was performed in PubMed, Embase, EBSCO, Ovid, Scopus, Web of Science, and Springer Link, focusing on observational and interventional studies using big data, public databases, EHR systems, wearables, and sensors for NCD management from January 2014 to July 2024. Data extraction included study characteristics, modeling approaches, and external validation, with synthesis through keyword analysis, the 3V framework, and visual tools such as word clouds, heat maps, and spider diagrams.

Results:

Results:

A total of 8,707 records were identified, and 90 studies were included for full-text review. The studies were categorized into four types of health portraits: diagnostic, prognostic, monitoring, and recommender. Data utilization based on the 3V framework showed that only 17.78% of studies met all 3V criteria. In terms of Volume, structured data was widely used (64.29%–100% depending on portrait type), while unstructured data showed substantial variation (19.05%–93.33%). Regarding Velocity, monitoring and recommender portraits showed high reliance on online interactive data (85%+). For Variety, only 31.11% of portraits utilized all three attributes. Comprehensive capability assessment revealed that only 30% of studies had external validation, and only 10% met both external validation and 3V criteria, with recommender portraits performing better in both areas.

Conclusions:

Conclusions:

The results highlight significant disparities in data utilization across portrait types and underscore critical challenges, including data continuity and reliability, limited cross-functional integration, and privacy risks, constraining their multidimensional utilization and external validation. Future research should focus on cost-effective cohort datasets from wearable and contactless devices and improve intervention and follow-up evidence to ensure reliability and effectiveness in real-world applications.