The method
1. Calculate before interpreting
Subtract the baseline score from the follow-up score. Check which direction represents improvement and whether the protocol stayed comparable. Arithmetic tells you the size of the observed change, not its cause.
2. Ask two different threshold questions
Minimal detectable change (MDC) concerns change relative to measurement error at a stated confidence level. Minimal clinically important difference (MCID) concerns importance. They are not interchangeable. Estimates must match the instrument, population and context.
3. Do not convert a threshold into a guarantee
A change can exceed one supplied threshold without exceeding the other. That calls for a careful statement, not a forced “success” or “failure” label. This fictional exercise is not a threshold for any real patient or instrument.
WORKED EXAMPLE
Fictional Learning Measure A
| Item | Supplied fictional value |
|---|---|
| Baseline | 40 |
| Follow-up | 46 |
| MDC95 | 8 |
| MCID | 5 |
Assume higher scores are better and both measurements used the same protocol. Baseline is 40 and follow-up is 46, so observed change is +6. The supplied fictional MDC95 is 8 and MCID is 5. Six exceeds five but not eight. Describe that distinction explicitly: the change exceeds the supplied importance threshold but does not exceed the supplied detectable-change threshold. This does not prove treatment caused the change or that it matters to a particular person.
PUT IT INTO PRACTICE
Your workshop
Your answers stay in this open page unless you choose to save on this device or download. Use fictional or de-identified learning examples; leave out patient and other people’s private information.
Check the reasoning
Choose an answer to see an explanation. These are learning checks, not an official exam or score prediction.
Build your worksheet
Try a changed situation
For a second fictional measure, baseline is 20, follow-up is 29, MDC95 is 7 and MCID is 10. Interpret only these supplied numbers.
Take away: Keep “how much?”, “beyond measurement error?” and “important?” as separate questions.
Answer both learning checks to enable your review record.
Marking reviewed records your own activity. It does not establish mastery or readiness.