Abstract
Introduction: The Fracture Risk Evaluation Model (FREM) utilizes national registry data to estimate an individual's risk of major osteoporotic fractures (MOF). FREM holds the potential to serve as a data-driven decision-support tool for general practitioners, thereby contributing to improve case-finding and treatment of osteoporosis patients.
Purpose: To investigate whether adding self-reported clinical data on known risk factors for osteoporosis to the FREM algorithm, improves precision in predicting 1-year osteoporotic fracture risk.
Methods: We obtained self-reported data on known risk factors for osteoporosis from women randomly selected to participate in the Danish ROSE (Risk-stratified Osteoporosis Strategy Evaluation) study who answered a questionnaire when invited to participate. Information on the risk factors in FREM’s underlying algorithm were obtained through individual record linkage with the national health registers.
We calculated the 1-year predicted risk of MOF using both FREM on its own and FREM combined with clinical factors for all individuals in the cohort and categorized individuals into risk groups (<2%, 2-3%, 3-4%, 4-5% and >5%). Observed MOF incidence in each of these predicted risk groups was compared to predicted risk separately for FREM and for FREM combined with clinical risk factors.
Results: Among the 18,605 Danish women included (median age of 70 years), 280 (1.5%) sustained MOFs within one year. Observed incidence of MOF aligned well with predictions from FREM alone and FREM combined with clinical factors for low-risk (<2% and 2-3%) and high-risk (>5%) individuals. However, both models slightly overestimated fracture risk for medium-risk individuals (3-4% and 4-5%).
Purpose: To investigate whether adding self-reported clinical data on known risk factors for osteoporosis to the FREM algorithm, improves precision in predicting 1-year osteoporotic fracture risk.
Methods: We obtained self-reported data on known risk factors for osteoporosis from women randomly selected to participate in the Danish ROSE (Risk-stratified Osteoporosis Strategy Evaluation) study who answered a questionnaire when invited to participate. Information on the risk factors in FREM’s underlying algorithm were obtained through individual record linkage with the national health registers.
We calculated the 1-year predicted risk of MOF using both FREM on its own and FREM combined with clinical factors for all individuals in the cohort and categorized individuals into risk groups (<2%, 2-3%, 3-4%, 4-5% and >5%). Observed MOF incidence in each of these predicted risk groups was compared to predicted risk separately for FREM and for FREM combined with clinical risk factors.
Results: Among the 18,605 Danish women included (median age of 70 years), 280 (1.5%) sustained MOFs within one year. Observed incidence of MOF aligned well with predictions from FREM alone and FREM combined with clinical factors for low-risk (<2% and 2-3%) and high-risk (>5%) individuals. However, both models slightly overestimated fracture risk for medium-risk individuals (3-4% and 4-5%).
Original language | English |
---|---|
Publication date | May 2024 |
Publication status | Published - May 2024 |
Event | European Calcified Tissue Society - Marseille, France Duration: 25. May 2024 → 28. May 2024 https://www.ects2024.org/ |
Conference
Conference | European Calcified Tissue Society |
---|---|
Country/Territory | France |
City | Marseille |
Period | 25/05/2024 → 28/05/2024 |
Internet address |