The study retrospectively assessed the medical records of women aged 65 and over who were prescribed monthly preparations of alendronate (21,615 patients) or risedronate (12,215 patients) between July 2002 and September 2004.
Fracture rates were compared for the first year of treatment between the two groups after adjustment for potential differences in baseline fracture risk. In total 109 hip fractures occurred, 29 in the risedronate group and 80 in the alendronate group. Statistical correction of the data produced percentage fracture rates of 0.37% of the risedronate arm and 0.58% of the alendronate arm.
The Numbers Needed to Treat (NNT) is 476. This means 476 patients would need to be treated with risedronate instead of alendronate for a period of one year to prevent one hip fracture. This would be at an approximate cost of £81,500. This also assumes that, as an observational study, there are no biases in the data affecting the conclusion.
Action: This study has produced an interesting hypothesis but should not change current practice without additional research. Alendronate continues to be the first line choice where a bisphosphonate is indicated.