A more evidence based approach to the use of surrogate end points in policy making

By retracing the history of particular surrogate outcome failures,1 Moynihan dismisses the assessment of surrogate outcomes in future clinical trials and instead recommends collecting only final outcomes such as survival.2 We agree that, when possible, decisions on the adoption and coverage of health technologies should be based on evidence of effectiveness and cost effectiveness derived from well conducted trials that report final relevant patient outcomes. Nevertheless we take issue with Moynihan’s dismissal of surrogate outcomes in policy making on at least two grounds.

First and foremost is the current context of technology licensing and the pressure for so called faster access to effective innovations—for example, the accelerated approval procedure of the US Food and Drug Administration. This will continue to present policy makers with market access and coverage decisions that may be based only on changes in surrogate outcomes.

Secondly, Moynihan fails to acknowledge the large body of methodological evidence already collected on the statistical validation of proposed surrogate outcomes,3 as well as the frameworks we and others have proposed to aid healthcare policy makers to take an evidence based approach in deciding whether they can trust a surrogate outcome.4 5

In conclusion, we encourage analysts and healthcare policy makers to make an evidence based judgment on the reliability of surrogate outcomes with reference to current validation frameworks.