Determination of risk threshold for osteoporosis therapy: a decision curve analysis approach — ASN Events
  1. Schedule
  2. New Investigator session
  3. Determination of risk threshold for osteoporosis therapy: a decision curve analysis approach

Determination of risk threshold for osteoporosis therapy: a decision curve analysis approach (#56)

Thao P. Ho-Le 1 , Jackie R. Center 2 3 , John A. Eisman 2 3 4 , Hung T. Nguyen 1 , Tuan V. Nguyen 1 2 3 4 5
  1. Centre for Health Technologies, Falcuty of Engineering and Information Technology, University of Technology, Sydney, Australia, Sydney, New South Wales, Australia
  2. Bone Biology Division, Garvan Institute of Medical Research, Sydney, NSW, Australia
  3. St Vincent Clinical School, University of NSW, Sydney, New South Wales, Australia
  4. School of Medicine, Notre Dame University , Sydney, New South Wales, Australia
  5. School of Public Health and Community Medicine, UNSW, Sydney, New South Wales, Australia

Despite of the availability of fracture prediction models, there is no scientifically validated threshold of fracture risk for the treatment of osteoporosis. In this study, we employed the decision curve analysis to derive the predicted probability of fracture that is optimally suitable for selecting individuals for osteoporosis treatment.

The study was part of the Dubbo Osteoporosis Epidemiology Study that involved 2188 women and 1324 men aged 60 years and above who have been followed up to 20 years. During the follow-up period, the incidence of fractures and mortality was ascertained. Baseline bone mineral density (BMD) and clinical risk factors were obtained prior to the fracture event. We considered 3 mortality-adjusted models for predicting fracture risk: Model I included age and femoral neck BMD; Model II included age, femoral neck BMD, history of fracture and falls; and Model III had factors in Model II plus lumbar spine BMD. For each model and each predicted probability of fracture, we assessed the ratio of true positives over false positives. The optimal threshold was defined as the point of highest ratio (i.e. net benefit). 

In women, compared with the strategy of treating everyone, treating those with 5-yr predicted fracture risk >10% yielded a better benefit. The 5-yr risk threshold that yielded the greatest benefit was ~20% for Model III. For hip fracture, Model II and Model III showed the same net benefit improvement at the 5-yr risk threshold of 4%. In men, adding lumbar spine BMD into Model II didn’t improve the net benefit. In men, the 5-yr risk of any fracture 10% or 5-yr risk of hip fracture 2% yielded the greatest benefit. 

Thus, we propose that individuals with 5-yr fracture risk of greater than 10% (for men) and greater than 20% (for women) should be considered "high risk" and may be indicated for treatment.