Prospective independent and external validation of QFractureScores: predicting risk of osteoporotic and hip fracture in the United Kingdom
Can we predict fractures?
Risk prediction models are useful tools for clinicians. They can identify people with a purported high risk of disease who might benefit from early intervention to reduce costly and avoidable complications of disease. Risk prediction models are useful only if they can be confidently applied to a large proportion of the population and if they target a disease for which there is evidence to support early intervention. When a new model is developed, it is important that the performance is tested in a fresh cohort that reflects the one in which it is likely to be used. This process is known as external validation.
The authors of this study wanted to externally validate new risk prediction scores, the QFractureScores,1 which were recently developed to help doctors predict the 10 year risk of osteoporotic fractures (vertebral, distal radius, or hip) and hip fractures in men and women. External validation is an