Surrogate outcome
Also known as: Surrogate endpoint, Proxy measure
A stand-in marker used instead of the outcome you actually care about.
What it means
A surrogate outcome is an intermediate measure substituted for a more meaningful but harder-to-observe endpoint, on the assumption that changing the surrogate implies changing the true outcome. Surrogates are attractive because they can be measured sooner, more cheaply, or more sensitively than endpoints like survival or real-world behavior. The danger is that a treatment can move the surrogate without improving the outcome that matters, or even worsen the latter while flattering the former. Validating a surrogate requires evidence that effects on it reliably predict effects on the true endpoint, a bar many widely used surrogates fail to clear.
Examples
Approving a drug because it lowers a cholesterol marker, when what patients ultimately care about is fewer heart attacks.
A support team is graded on average call-handling time, the stand-in for happy customers. Calls get shorter, the dashboard glows green, and repeat complaints quietly climb.
A district rewards teachers for raising test scores, the surrogate for learning. Scores rise on drilling and exam tactics while reading anything beyond the test does not.
First described in Clinical-trials methodology; Prentice criteria (1989).