Predictive Data Analytics in Telecare and Telehealth: Systematic Scoping Review.

Area of Science:

• Digital Health
• Health Informatics
• Data Science

Background:

• Telecare and telehealth services support independent living at home.
• Historically reactive, these technologies are evolving towards proactive and predictive care.
• A recent drive focuses on leveraging data for enhanced predictive capabilities in home-based care.

Purpose of the Study:

• To review the application of predictive data analytics in telecare and telehealth for at-home settings.
• To explore current trends and identify future opportunities in this domain.

Main Methods:

• Systematic scoping review adhering to PRISMA-ScR guidelines and Arksey and O'Malley's framework.
• Searched MEDLINE, Embase, and Social Science Premium Collection (2012-2022).
• Included English language papers, screened against defined inclusion/exclusion criteria.

Main Results:

• 86 papers were included, categorizing analytics into anomaly detection (21), diagnosis (32), prediction (22), and activity recognition (11).
• Parkinson disease (12) and cardiovascular conditions (11) were the most studied health conditions.
• Key findings include underutilization of routinely collected data, a prevalence of diagnostic tools, and identified barriers/opportunities for future predictive analytics.

Conclusions:

• Current applications are predominantly small-scale pilots; larger trials are needed for predictive techniques.
• Integrating routinely collected care data and patient-reported outcomes can significantly enhance predictive models.
• Future research must ensure data sets are sufficiently large and diverse for generalizable and robust model training, validation, and testing.