Prediction tools based on electronic health records, artificial intelligence/machine learning, and wearable devices for identifying older adults at risk of falls or fall-related fractures: protocol for a systematic review and evidence synthesis

This project describes a systematic review and evidence synthesis aimed at evaluating prediction tools based on electronic health records, artificial intelligence and machine learning techniques, and wearable sensor technologies for identifying older adults at risk of falls or fall-related fractures. The review will include the development and validation of prediction models using routinely collected clinical data and sensor-derived digital biomarkers of gait, balance, and mobility. The primary objectives are to assess the diagnostic and prognostic accuracy, calibration, and validation of these tools, as well as their methodological quality and risk of bias. Where appropriate, quantitative synthesis of performance metrics will be conducted. By providing a comprehensive and up-to-date overview of emerging digital prediction tools, this review aims to support clinicians, researchers, and policymakers in understanding their potential role and readiness for real-world implementation in fall prevention strategies for ageing populations.