Analysis of Near-miss Data at the Aged Care Facility and Suggestions for Improvement

Facing a super-aged society, Japan experiences increasing demand for efficient senior care. This study addresses the underutilization of near-miss reports, valuable for accident prevention, in aged care facilities. Researchers conducted interviews at a care facility in Japan and identified key issues in recordkeeping, particularly that internal reports are rarely used for feedback or improvement. Focusing on 407 near-miss cases over six months, the study applied text mining, clustering, decision tree learning, and co-occurrence network analysis. Key findings include that “going to the toilet alone” and incidents in the dining room were major factors to near-misses, often involving specific individuals. Seasonal and individual fluctuations were also observed. The results were shared with care staff, leading to actionable measures such as targeted monitoring and consideration of automated alerts for fall risks. The study concludes that data-driven approaches can significantly improve safety and efficiency in elder care.