MCID, MDC and Measurement Error: Interpreting Change in Rehabilitation Outcomes
A clinician-friendly guide to deciding whether a change in an outcome measure is real, important, both or neither.
What You Should Be Able to Explain After Reading
- Differentiate reliability, measurement error, MDC and MCID.
- Explain why statistical change and patient-important change are different questions.
- Interpret thresholds in the population and context in which they were derived.
- Avoid treating published cutoffs as universal constants.
Start with the question
When an outcome score changes, clinicians usually want to know two things: is the change larger than expected measurement noise, and is it important to the patient? These are related but different questions.
MDC and measurement error
The minimal detectable change is derived from measurement error and estimates how large a change must be before it is unlikely to reflect error alone at a specified confidence level. It is a measurement-property concept, not a direct statement of patient importance.
MCID or minimal important change
A minimal important change attempts to represent a change considered meaningful, often using an external anchor such as patient global rating. Its value depends on population, baseline severity, anchor and method.
Clinical reasoning
- Use the same instrument and scoring method at comparable time points.
- Check whether the published threshold comes from a population similar to your patient.
- Interpret the score alongside the patient's stated goals and objective function.
- Remember that group-level research thresholds can be uncertain when applied to one individual.
Worked interpretation
If a patient's score improves beyond an estimated MDC but not an established important-change threshold, the change may be larger than measurement noise without yet being clearly important. If it exceeds an important threshold but remains within large measurement error, confidence in the individual change is limited.
Common mistakes
- Calling every statistically significant difference clinically important.
- Using an MCID from another diagnosis without justification.
- Assuming one cutoff is universal across baseline severity and settings.
- Ignoring direction of scoring and instrument version.
Clinical pearls
- MDC asks about confidence beyond error; MCID asks about importance.
- Thresholds support clinical reasoning; they do not replace it.
- Always pair outcome scores with the patient's actual functional story.
Summary
Interpret change by combining measurement precision, evidence about meaningful change, the correct population and the patient's own functional goals.
Evidence & Further Reading
These sources are provided so learners can trace the clinical recommendations and read the original guidance.