Conclusions about intervention effects should not be based on surrogate end points : [Letter]

Experts of controlled clinical trials argue that decisions on medical interventions should be based on clinically relevant outcomes and not on surrogates such as laboratory measurements. There are quite a few examples in which the effect on a surrogate end point substantially diverged from the effect on a clinically relevant outcome. In this respect, the recent paper by Bruno et al. is problematic as it proposed higher vitamin E intakes for smokers on the basis of greater disappearance rate of alpha-tocopherol in the plasma of smokers. The disappearance rate is a surrogate end point with no validated relation to any clinically relevant outcome.Non peer reviewe