[Surrogate end points--somewhere between essential and not worth discussing]

Abstract

Surrogate markers are only acceptable in single exceptional cases to demonstrate a patient related true outcome in medical care. This is especially true in drug therapy. Although many surrogate markers are correlated with an outcome, only a few have been shown to capture the effect of a treatment on the outcome. The examples for aberrations in medical care are numerous: They include menopausal hormone therapy, the prevention of cancer or cardiovascular diseases with vitamins, the therapy of cardiac arrhythmias using a special type of drugs as well as the treatment of osteoporosis with fluoride. All these treatments caused more harm than good to the patients; and very often mortality increased. In summary, most of the surrogate markers are not worth discussing, and they should not be utilised to demonstrate patient related outcome; they may only be acceptable in rare cases and would then have to be rigorously validated to protect patients from harm.