Glossary · Outcomes and measurementUpdated September 28, 2026

What is the minimal clinically important difference (MCID)?

The MCID is the smallest change in an outcome score that patients would perceive as beneficial and that would justify a change in their care. It turns a statistically significant result into a clinically interpretable one.

  • MCID in plain language
  • How it works in practice
  • Related terms

Free sandbox · No credit card · 21 CFR Part 11 aligned

Responder analysis · DLQI

Change from baseline

7 / 15

At or above MCID (4)
0-3 Below MCID4-15 At or above MCID (4)
Commonly cited DLQI MCID: 4 points

In short

  • Anchor-based methods link score changes to an external judgement, such as patients rating themselves "a little better". Distribution-based methods use statistical properties such as half a standard deviation.
  • MCIDs are instrument- and population-specific. Commonly cited examples: about 5 points on the PHQ-9, about 4 points on the DLQI.
  • Trials use MCIDs to define responders, to choose the effect size for sample size calculations and to interpret mean differences.
  • Regulators increasingly talk about meaningful within-patient change for patient-reported endpoints, estimated with anchor-based methods.

Explained

Statistically significant is not the same as meaningful

A large trial can detect a one-point improvement on a 30-point scale with a tiny p-value, even if no patient would notice the difference. The MCID provides the missing reference point: if the average improvement is well below the MCID, the result may not matter to patients, whatever its significance.

Because MCIDs depend on the population, baseline severity and method, published values vary. Use values derived in populations like yours, pre-specify the threshold in the analysis plan, and consider reporting the proportion of participants who reach it alongside mean differences. For sample size, see sample size calculation.

FAQ

Common questions

Something not covered here? Ask us directly.

What does MCID stand for?

Minimal clinically important difference. MCII (minimal clinically important improvement) is a related term.

How is an MCID calculated?

Anchor-based methods relate score changes to an external indicator of meaningful change; distribution-based methods use statistical properties of the scores. Anchor-based methods are generally preferred.

Is there one MCID per instrument?

No. MCIDs vary by population, severity and method, so use values from comparable populations.

How are MCIDs used in trials?

To define responders, choose effect sizes for sample size and interpret whether results are meaningful to patients.

Collect outcomes you can interpret

Validated instruments with coded answers. Free sandbox.

Try Capture free