Chang-Chun Chen, Alisha H Johnson, Marjorie Skubic, Shu-Fen Wung
This protocol provides an approach for evaluating contactless fall detection systems in long-term care environments. Future studies are needed to apply the protocol across different sensor platforms and care settings to evaluate reproducibility, generalizability, and operational feasibility.
BACKGROUND: Falls among residents of long-term care facilities are associated with substantial morbidity, mortality, and healthcare utilization. Despite the availability of numerous fall detection technologies, practical protocols for evaluating performance in real-world care environments remain limited.
OBJECTIVE: To develop a pragmatic and replicable protocol for validating contactless fall detection systems in long-term care environments.
METHODS: This article presents a protocol that combines structured testing by trained stunt actors, continuous real-world monitoring, and predefined procedures for event verification and classification to evaluate system performance across fall and non-fall events relevant to older adults.
PROTOCOL DESCRIPTION: The protocol consists of three phases: pre-implementation assessment, system implementation and training, and system verification. It incorporates site assessment, sensor deployment, scripted testing, continuous monitoring, event verification, and predefined analytical procedures under real-world conditions.
CONCLUSION: This protocol provides an approach for evaluating contactless fall detection systems in long-term care environments. Future studies are needed to apply the protocol across different sensor platforms and care settings to evaluate reproducibility, generalizability, and operational feasibility.