Digital Predictive Technologies for Fall Risk Assessment, Prediction, and Prevention in Long-Term Care: A Scoping Review.

OBJECTIVES: This scoping review aimed to map the available evidence on digital predictive technologies for fall risk assessment, prediction, and prevention support among older adults in long-term care settings. DESIGN: Scoping review. SETTING AND PARTICIPANTS: Included studies focused on older adults or residents in long-term care settings, including nursing homes, residential aged care facilities, assisted living facilities, skilled nursing facilities, and other institutional or residential long-term care environments. METHODS: A comprehensive search was conducted in PubMed, Scopus, Web of Science, and CINAHL for English-language original studies published from 2016 to 2026. Additional records were identified through manual reference searching and citation tracking. Eligible studies examined digital predictive technologies, including artificial intelligence, machine learning, deep learning, wearable sensors, inertial measurement units, passive sensors, Internet of Things-based systems, electronic health record-based prediction models, Minimum Data Set-based analytics, predictive dashboards, and clinical decision-support systems. Data were synthesized descriptively and narratively following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews guidance. RESULTS: A total of 2030 records were identified from electronic databases, and 15 additional records were identified through manual reference searching and citation tracking. After 845 duplicates were removed, 1185 database records were screened, and 60 reports were assessed for eligibility. Eleven studies met the eligibility criteria. The included studies used diverse technologies and data sources, including electronic health records, Minimum Data Set records, medication data, functional assessments, fall history, vital signs, wearable sensors, inertial sensors, passive sensors, and dashboard-linked data streams. Prediction windows varied from daily fall probability to 90-day, 3-month, and 6-month fall risk prediction. Reported outcomes included fall risk probability, future falls, recurrent falls, major falls, and fall rates. Model performance varied across studies, with evidence ranging from proof-of-concept findings to moderate or good discrimination. However, most studies focused on model development, validation, feasibility, or early implementation, and only limited evidence directly evaluated fall reduction or sustained clinical outcomes. CONCLUSIONS AND IMPLICATIONS: Digital predictive technologies show emerging potential to support fall risk assessment and prevention planning in long-term care. However, the evidence base remains small, heterogeneous, and largely developmental. These technologies should be viewed as decision-support tools that complement, rather than replace, clinical judgment and nursing assessment. Future research should prioritize prospective validation, workflow integration, ethical governance, cost effectiveness, staff response, resident-centered outcomes, and real-world effects on fall reduction in diverse long-term care settings.