Vertebral fracture AI detection for better osteoporosis care (the VertAIdo-project)

Purpose and goal: Aim:

Investigate whether opportunistic AI screening (vertebral fractures) integrated with fracture liaison service leads to a higher number of correctly diagnosed and treated patients and whether it is health-economically beneficial compared to present care. Objectives include:

Does an AI-support implemented in a clinical routine flow lead to a higher number of vertebral fractures being diagnosed, patients investigated and treated compared to ordinary work methods? Expected results and effects: Hypothesis:

Opportunistisk AI-detektion of vertebral fractures that is well integrated with the fracture chain will significantly increase the number of diagnosed and treated vertebral fractures and is cost-effective compared to current systems. Finding and treating these patients will lead to a lower risk of new serious osteoporosis fractures.

Approach and implementation:

The study will compare 4 months of AI support implemented in a clinical fracture chain (intervention) compared to a historical cohort of the same time period the year before (control). The study cohort includes patients >50-years of age undergoing CT thorax and/or abdomen, regardless of cause. The actual prevalence of vertebral fractures on CT examinations is expected to be constant during this time.