In vitro stress shielding measurements can be affected by large errors.

Hip prostheses and other implantable devices for the proximal femur are tested experimentally to study their effects on load transfer. We report on some experimental errors (related to the load simulation) that can undermine the reliability of strain measurements. A first source of error is that of overconstraining the setup. This situation makes it impossible to control or even determine the force values. The second source of error is related to geometric alterations induced by surgery, which modify the lever arms and thus the loading system. Two options are available to compensate for a geometric alteration: either applying the same system of forces or the same resultant bending moment to the implanted femur. The errors that arise if these parameters are not controlled can make it impossible to determine if one device performs better or worse than another.