Predicting Fracture in Long-Term Care

Aim: To explore which factors identified from RAI-MDS (Resident Assessment Instrument – Minimum Data Set 2.0) are associated with incident fracture in one year along long-term care residents and to use these analyses to inform the development and testing of one or more fracture risk prediction algorithms.

Overview: Frail older adults living in LTC homes face many challenges including a high risk of fracture. Hip fractures are common in the long-term care setting with over 1/4 of hip fractures occurring among residents. Fracture prediction tools available have been well characterized in community dwelling populations.  However, they have not been validated in LTC and use bone mineral density testing which may be challenging to obtain in LTC residents. These tools are based on 10-year fracture risk; however, 20% of residents may die within one year. In addition, these tools are missing LTC risk factors. 

Study Components:

  1. Analysis of variables associated with incident fractures using data from RAI-MDS 
  2. Decision-tree analysis using variables from RAI-MDS, DADS and NACRS (Discharge Abstract Database and National Ambulatory Care Reporting System) to develop fracture risk prediction algorithms. The analysis provided a number of different groups of at risk individuals with a similar probability of having a fracture
  3. Testing and validation of algorithm(s) created in #2
  4. Development of tool
  5. Review of tool by InterRAI Instrument Development Committee
  6. Global adoption of tool by all licensed LTC vendors

Results: A 1 year hip fracture prediction tool for LTC using electronic RAI-MDS data has been developed

Components of Success:

  • A tool to predict incidence of hip fracture in frail older adults in long-term care
  • The tool is effective at discriminating and predicting hip fractures in LTC residents, without using bone mineral density as a factor
  • The tool may be used as a Clinical Assessment Protocol for evaluating fracture risk residents
  • International adoption by all licensed vendors in long-term care and its equivalents (e.g. home care) is expected