To the Editor.
—As Dr Karpfand colleagues1 demonstrate, alendronate use is associated with a reduced rate of nonvertebral osteoporotic fractures in selected populations. The relative risk and confidence intervals are impressive but are not adequate to formulate clinical decisions. Practicing clinicians should be hesitant to translate the study data directly to the prescription pad.The number needed to treat (NNT),2 by its inverse relationship with absolute risk, has a much broader perspective on an intervention's effectiveness and provides an effort-yield ratio that a practicing clinician can use. It is a measure of how many people should receive an intervention to achieve the desired clinical end point in 1 person.From the alendronate data, the estimated cumulative incidence of nonvertebral fractures after 3 years was 12.6% in the control group and 9.0% in the alendronate group. From this the absolute risk reduction can be calculated to be 3.6%, and its inverse, the
Prostko M. Meta-analysis of Prevention of Nonvertebral Fractures by Alendronate. JAMA. 1997;278(8):631. doi:10.1001/jama.1997.03550080041022