Background: Several national organizations recommend that fracture risk assessment and osteoporotic treatment be based on estimated absolute 10-year fracture risk rather than bone mineral density (BMD) alone.
Objective: To assess the changes in physician prescribing behavior after introduction of absolute 10-year fracture risk reporting.
Design: Before-and-after study.
Setting: Manitoba, Canada, which has an integrated BMD program in which tests are linkable to a population-based administrative health database repository.
Patients: Women 50 years or older who were not receiving osteoporosis medication (2042 before and 3889 after intervention).
Intervention: Introduction of a system reporting absolute 10-year fracture risk along with dual-energy x-ray absorptiometry results.
Measurements: The proportion of untreated women who were prescribed osteoporosis medications in the year after baseline BMD measurement.
Results: Absolute fracture risk reporting reclassified more women (32.7%) into lower-risk categories than into higher-risk categories (10%). This effect was more prominent in women younger than 65 years. Fewer women per physician were prescribed osteoporosis drugs after introduction of absolute fracture risk reporting. The absolute fracture risk reporting system was associated with an overall reduction in osteoporosis medications dispensed (adjusted absolute reduction, 9.0 percentage points [95% CI, 3.9 to 14.2 percentage points]; relative reduction, 21.3% [CI, 9.2% to 33.5%]; P < 0.001). The reduction was attributed to fewer drugs dispensed to women at low and moderate risk for fracture. No differences in fracture rates were observed.
Limitations: This was a nonrandomized study. The risk assessment system studied differs slightly from other 10-year fracture risk assessment models.
Conclusion: Change from a T-score-based fracture risk reporting system to a system based on absolute 10-year fracture risk was associated with appropriate, guideline-based changes in prescription of osteoporosis medications.
Primary funding source: None.