Lead-time bias

In one sentence

Lead-time bias occurs when earlier detection lengthens measured survival from diagnosis without changing when the outcome occurs.

The intuition

Move the starting line of a stopwatch earlier while keeping its finish fixed. The measured interval grows, even though nothing changed at the finish. That is why finding disease sooner and improving survival are different achievements.

How it works

A molecular assay can detect a signal before imaging identifies recurrence. The interval between those detections is lead time. Its size depends on biology, assay sensitivity, draw frequency and imaging schedule. It is a cohort observation, not a guaranteed window in which every lesion remains small and treatable.

For a hypothetical timeline, recurrence is clinically diagnosed at year four and death occurs at year six. Measured survival after recurrence is two years. An assay diagnoses recurrence at year three, but death still occurs at year six. Measured survival is now three years without postponing death.

A strategy can overcome this concern by showing improvement in a meaningful outcome measured from an appropriate common origin, ideally with a suitable comparator. The benefit needs to be distinguished from extra time knowing about the disease, changes in labeling and selection of slower-growing cases.

Earlier detection can also trigger burdens: additional imaging, biopsies, anxiety or treatment exposure. Those do not make detection worthless; they belong in the utility question. A ctDNA (circulating tumor DNA)-positive result without a visible lesion is a different situation from a safely localizable recurrence. A local-treatment eligibility threshold cannot turn lead time into proof of benefit.

Why it matters in cancer

It distinguishes earlier detection from evidence that a monitoring strategy improves outcomes.

Worked example

An assay reports a median ten-month lead over scans. That does not establish ten months of safe delay, ten months of benefit, or an automatic opportunity for ablation. A trial of a defined action must answer whether earlier intervention helps.

Common confusions

Lead time is not added life.
A longer lead does not guarantee a localizable lesion.
Comparing survival from different diagnosis times can favor earlier detection without a true outcome change.

Endpoints and local-treatment timing.

Sources and scope

General teaching, source-checked October 8, 2026. The examples are hypothetical unless a study is explicitly named. Expert and learner review remain pending.