- FRACTURE-ML is a new AI tool developed from Swedish registry data that predicts hip fracture risk more effectively than current screening methods.
- The tool identified nearly seven times more high-risk individuals than current clinical practices, with a sensitivity of 0.84 compared to 0.12.
- The study analyzed data from 3,542,647 individuals aged 50+ who had not been prescribed osteoporosis medication in the prior two years.
- During the follow-up period of up to 10 years, 142,327 participants sustained a hip fracture.
- FRACTURE-ML achieved a one-year AUC of 0.89 and a five-year AUC of 0.85, indicating high predictive accuracy.
- A simplified version of the model, using 35 variables, performed nearly as well with a one-year AUC of 0.88.
- The findings were published on August 27 in the journal PLOS Medicine.
- Hip fractures are common in older adults and can lead to serious consequences regarding independence and health.
- Existing tools for predicting fracture risk typically require patient-provided information, complicating large-scale screening.
- FRACTURE-ML could shift hip-fracture care from reactive treatment to preventive measures.
A new AI tool, FRACTURE-ML, developed by researchers at the University of Gothenburg, predicts hip fracture risk with remarkable accuracy using data from Sweden's national health registers.110
The model analyzed data from 3,542,647 individuals aged 50 and older, identifying 142,327 hip fractures during a follow-up period of up to 10 years.48
Using over 100,000 variables related to health and demographics, FRACTURE-ML achieved a one-year area under the curve (AUC) of 0.89 and maintained an AUC of 0.85 at five years.
Compared to current screening methods, it identified nearly seven times more high-risk individuals within two years, with a sensitivity of 0.84 versus 0.12 for existing practices.2
“The findings show that it is possible to predict hip fracture risk at the population level without direct patient interaction,” said Kristian Axelsson, MD. “This approach could help target preventive measures more efficiently and potentially reduce the number of hip fractures.”
The tool's reliance on existing registry data allows for large-scale screening without the need for patient assessments, making it a promising solution for preventive care.
“Hip fractures often result in significant suffering, loss of independence, and increased mortality,” noted Mattias Lorentzon, MD. “Our model is very good at distinguishing between individuals at high and low risk and has the potential to become an important tool for preventive care.”
The study, published in PLOS Medicine, highlights the potential of FRACTURE-ML to transform hip fracture prevention strategies.7
“The tool, developed from Swedish registry data of over 3.5 million adults, achieved an AUC of 0.89 at one year and 0.85 at five years. A simplified 35-variable version performed nearly as well, suggesting practical clinical use.”
News-Medical
