That is what a reader asked.
We can determine which Antiresorptive medicine is “best” for an individual based on avoiding likely problems unique to that individual. We cannot yet determine which medicine is “best” for preventing that first Osteoporotic Fragility Fracture in everyone.
At first glance, it appears that the FDA has determined that each medication works in the specific areas of Vertebral (Spine), Non-Vertebral, or Hip. But if you look at the data, it is more complicated.
First, there are no large comparisons of one med to another. There are comparisons to a “control” group of patients (who are similar to the patients taking the med being studied, but do not take the med).
Second, the medical and fracture risk details of each study group (both those taking the med and those not) are different from the details of every other study group. Therefore, you cannot make valid comparisons among different meds based on the effectiveness of each med vs. its own unique “control” group.
Third, results are considered “real” based on the number of patients studied as well as the numbers of fractures in the “treatment” and “control” groups. If the statistical calculation determines a 5% or less likelihood that the difference in fracture numbers is due to random chance alone, the difference is considered “real”.
The more patients in the study, and the more fractures in the “control” group, the easier it is to “prove” the med is “effective”. All meds had large enough study groups to “prove” effectiveness in Vertebral and Non-Vertebral fracture prevention. Hip fractures are relatively rare, so numbers are often too small.
Actonel statistics showed a 4.9% likelihood of random chance. Boniva showed 5.1%. One more hip fracture in the control group for Boniva and one less in the contol group for Actonel would have reversed the results. The FDA stuck to the 5% rule. But is there really a statistical difference when the study groups were not the same?
Different classes of meds are next.
Jay Ginther, MD