Surrogate Outcomes in Trials: A Number Versus a Patient Benefit

A surrogate outcome is a measurement a trial uses in place of what patients care about most: how they feel, how well they function, or whether they live longer. A lab number or scan result can move in the right direction without proving that people are better off. Before you trust a trial result, check which kind of outcome was measured and how strong the proven link is between that number and a real patient benefit.

By ClinicalStudyConnect.com Research Desk

Two Kinds of Outcomes, in Plain Language

An outcome (also called an endpoint) is what a trial measures to see whether something worked. The FDA-NIH BEST glossary, a shared dictionary of research terms, separates two kinds:

Clinical outcome: “An outcome that describes or reflects how an individual feels, functions or survives.”
Surrogate endpoint: “An endpoint that is used in clinical trials as a substitute for a direct measure of how a patient feels, functions, or survives.”

Many surrogates are biomarkers. BEST defines a biomarker as “a defined characteristic that is measured as an indicator of normal biological processes, pathogenic processes, or biological responses to an exposure or intervention.” In everyday terms, a biomarker is something you can measure in the body, like a blood test value or a blood pressure reading.

Surrogates are not bad science. The U.S. Food and Drug Administration (FDA) explains that they are used when measuring clinical outcomes would take too long, or when the link to clinical benefit is already well understood. The key question is how well a given surrogate has been shown to predict real benefit.

The Evidence Ladder: How Strong Is the Link to Real Benefit?

Use this ladder to place any trial outcome you read about. The top four rungs come from the FDA and BEST definitions. The bottom rung is our own reading tip. Higher rungs give you more direct evidence that people actually benefit.

Clinical outcome. The trial directly measured how people felt, functioned, or survived.
Validated surrogate endpoint. BEST describes this as an endpoint “supported by a clear mechanistic rationale and clinical data providing strong evidence that an effect on the surrogate endpoint predicts a specific clinical benefit.” The FDA gives an example: across multiple trials, lowering systolic blood pressure (the top number) has predicted fewer strokes.
Reasonably likely surrogate endpoint. Backed by a strong biological or population-study reason to expect benefit, but, in the FDA’s words, “the amount of clinical data available is not sufficient to show that they are a validated surrogate endpoint.” The FDA notes these can support its Accelerated Approval program for serious diseases.
Candidate surrogate endpoint. BEST calls this an endpoint “still under evaluation for its ability to predict clinical benefit.” The link to patient benefit is not established yet.
Our reading tip: a measurement with no stated link. If a study reports a change in a number and does not explain what evidence connects it to how people feel, function, or survive, treat the finding as early and unconfirmed.

Where does an intermediate clinical endpoint fit? BEST describes it as a clinical outcome that can be measured earlier than permanent harm or death, and that is “considered reasonably likely to predict” longer-term benefit. It is a real clinical measure, but it is still a step before the final outcome.

Known Versus Unknown: What a Surrogate Result Can and Cannot Tell You

What a surrogate result can show

That the measured number changed in the group studied, over the time period studied.
That the change may point toward a benefit, depending on which rung of the ladder the surrogate sits on.
For a validated surrogate, that strong prior evidence connects that number to a specific clinical benefit.

What a surrogate result cannot show by itself

That people in the trial felt better, functioned better, or lived longer. That was not what was measured.
That a link shown for one treatment carries over to a different treatment, ingredient, or group of people. The BEST definition ties validation to predicting “a specific clinical benefit.”
The full picture of harms. A trial focused on a number may not be designed to capture all side effects or long-term effects.
That a candidate or reasonably likely surrogate will turn out to predict benefit once more clinical data exists.

How Reporting Standards Help You Spot the Outcome That Matters

Two international guidelines set out what trial reports and trial plans should include:

CONSORT (Consolidated Standards of Reporting Trials) is described by its developers as “an evidence-based, minimum set of recommendations for reporting the results of randomised trials.” The current version, CONSORT 2025, has a 30-item checklist.
SPIRIT (Standard Protocol Items: Recommendations for Interventional Trials) is “a minimum set of recommendations for reporting the protocol of a randomised trial.” A protocol is the trial’s written plan. The current version, SPIRIT 2025, has a 34-item checklist.

For surrogate outcomes, the most useful CONSORT 2025 reporting topics are:

Defining the primary outcome (the trial’s main outcome) and secondary outcomes, including how and when they were assessed.
Reporting important changes to the trial after it started, including any outcomes or analyses that were not prespecified (planned in advance).
Explaining how harms and other unintended effects were assessed.
Saying where the trial was registered and where the protocol and statistical analysis plan can be accessed.
Disclosing the authors’ financial and other conflicts of interest.
Discussing the trial’s limitations, including how widely the results may apply.

If a report is missing these details, that does not prove the trial was wrong. It does mean you have less information to judge whether the outcome reflects a real patient benefit.

Your Next-Step Checklist

Use this checklist with any trial summary, abstract, or registry record. You do not need a science background to work through it.

Find the primary outcome. Look in the abstract or methods section, or in the trial’s registry record, such as its entry on ClinicalTrials.gov. Write down exactly what was measured.
Sort it: clinical outcome or surrogate? Ask: does this directly describe how people felt, functioned, or survived? If not, it is a surrogate or another intermediate measure.
Place it on the evidence ladder. Does the report explain whether this surrogate is validated, reasonably likely, or still a candidate? If it says nothing, treat the link as unconfirmed.
Check whether the outcome was planned. Compare the reported primary outcome with the registry record or protocol. Note any changes and whether the authors explained them.
Check the time frame. A short trial may show a change in a number long before any clinical outcome could appear.
Look for harms. See how side effects and unintended effects were measured and reported, not just whether the target number improved.
Check who was studied. A link shown in one group of people may not apply to people of different ages, health status, or conditions.
Note funding and conflicts. Record who paid for the trial and any author conflicts disclosed.
Read the limitations. Authors should say what their trial could not show. Take that section seriously.

Questions to Bring to Your Doctor or Pharmacist

If a trial result is part of a decision about your own health, these neutral questions can help you have a clearer conversation:

Did this study measure a lab value or test result, or how people actually felt, functioned, or lived?
Is this measurement considered a validated predictor of the benefit I care about?
Were the people in this study similar to me?
What did the study find about side effects, and how long did it follow people?
Is there newer or larger research on clinical outcomes for this question?

Do not start, stop, or change any medicine or supplement based on a single study result. Talk with a qualified health professional who knows your history.

Sources

About This Guide

This article is general educational information about how clinical trials are designed and reported. It is not medical advice. It does not diagnose any condition, recommend any treatment, or evaluate any product. If you think you are having a medical emergency, contact your local emergency services right away. ClinicalStudyConnect.com is an independent educational publication and does not sell or promote any product.