Novel prediction model may accurately assess 5-year hip fracture risk
Click Here to Manage Email Alerts
Key takeaways:
- A novel clinical model may accurately predict risk of hip fracture without data on bone density.
- The model accurately predicted the 5-year hip fracture risk among the Swedish population.
A novel clinical prediction model accurately assessed the 5-year hip fracture risk among the Swedish population without data on bone density, according to published results.
Peter Nordström, MD, from the department of public health and caring sciences at Uppsala University in Sweden, and colleagues used cross-linked data from nationwide registers in Sweden to perform an observational study of all Swedish residents aged 50 years or older in 2005 (n = 3,340,977).
According to the study, Nordström and colleagues used a novel clinical model that assessed 19 variables, including diagnoses, prescription medications, ancestry, frailty and socioeconomic factors, to predict 5-year hip fracture risk among the cohort.
With a total follow-up of 15.2 million person-years among all patients, Nordström and colleagues found 87,089 people had a hip fracture within 5 years at a mean age of 82.4 years. They found the prediction model was associated with a population attributable fraction for hip fracture of 92.7% in men and 93.9% in women.
Nordström and colleagues found the use of home health care services was the strongest predictor of hip fracture, with a 5-year risk of 4.7% in men and 7.8% in women. They hypothesized that use of home health care services may be associated with frailty.
Among diagnoses, Nordström and colleagues found Parkinson’s disease (4.6% in men; 6.8% in women) and dementia (3.6% in men; 6.1% in women) were the strongest predictors of hip fracture.
According to the final validated prediction model, people with an estimated 5-year risk for hip fracture of at least 3% should be considered for bone-specific treatment. Using a treatment threshold of 3%, Nordström and colleagues found a 30% relative risk reduction after bone-specific treatment. They found the number needed to treat to prevent one hip fracture was 36 in women and 52 in men.
After performing subanalysis of 504,431 non-Swedish residents, Nordström and colleagues found comparable outcomes vs. the Swedish cohort.
“Given the lack of equipment to measure bone density, and the fact that most individuals with fractures do not have osteoporosis, it would be important if fracture risk could be determined without measuring bone density,” Nordström and colleagues wrote in the study.