Fracture risk assessment in the presence of competing risk of death
Authors' conclusion
Adjustment for competing risk of death has minimum impact on the short-term prediction of fracture. However, the multistate model yields the most accurate prediction of long-term fracture risk
Study context
- Population
- Elderly participants of the Dubbo Osteoporosis Epidemiology Study monitored for bone health and mortality
- Participants
- 3,035 participants
- Comparison
- Four fracture-prediction models: conventional Cox, cause-specific, Fine-Gray sub-distribution and multistate
- Outcome
- Prediction of fragility fracture risk in the presence of competing death (discrimination and calibration)
Methodological detailsdevelopment (60%) and validation (40%) subsets, discrimination and calibration analyses
- Development (60%) and validation (40%) subsets
- Discrimination and calibration analyses
Performance
As reported in the abstract
10-year fracture risk prediction, calibration-in-the-large index · 10 years
Multistate
- 0.05
95% CI - 0.20, 0.10
Cause-specific
- 0.23
95% CI - 0.30, - 0.08
Fine-Gray
- 0.31
95% CI - 0.46, - 0.16
Conventional Cox
- 0.54
95% CI - 0.70, - 0.39
Further findings
| Outcome | Values |
|---|---|
| Calibration-in-the-large index, 5-year fracture risk prediction | Conventional - 0.24 (95% CI - 0.43, - 0.06) |
Calibration-in-the-large index, 5-year fracture risk prediction
- Neither the discrimination nor the 5-year prediction performance was significantly different among the models; the conventional model tended to overestimate fracture risk
Original publication
Fracture risk assessment in the presence of competing risk of death. Osteoporos Int. Published 15 August 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →