Key points
- A change only matters if it's larger than the test's own measurement noise.
- For older adults, published estimates put a small meaningful change in gait speed at about 0.05 m/s and a substantial change at about 0.10 m/s.
- Standardized capture and trends across several visits make real change easier to see than any single before-and-after pair.
The question behind every re-test
A patient walks 0.03 m/s faster than last month. Their Timed Up and Go is a second quicker. Is that progress, or just a different day? Every re-assessment raises the same question, and the honest answer depends on two things: how large the change is, and how much the measurement would have moved anyway.
Outcome measures are only useful for tracking if clinicians can separate genuine change from the ordinary variability built into any test — differences in footwear, fatigue, instructions, timing, or who is holding the stopwatch.
Two thresholds worth knowing
Research on outcome measures generally describes change in two ways. A measurement-error threshold asks whether a difference is larger than the test's own noise. A meaningful-change threshold asks whether the difference matters to the patient's function.
For common performance measures in older adults, published estimates offer useful anchors. A widely cited analysis by Perera and colleagues (2006) placed a small meaningful change in gait speed at roughly 0.05 m/s and a substantial change at roughly 0.10 m/s. The same work estimated small and substantial change at about 0.5 and 1.0 points on the Short Physical Performance Battery, and about 20 m and 50 m in six-minute walk distance. We've compiled these, with citations, in our functional test reference values.
These are population estimates, not rules for any individual. They're a starting point for interpretation, not a replacement for clinical judgment.
Why noisy measurement erases real progress
Thresholds only help if the measurement is consistent enough to detect them. If a hand-timed walk varies by a tenth of a meter per second from one session to the next simply because of how it was run, a genuine 0.05 m/s improvement is invisible — and a spurious "decline" can look like a reason to change the plan of care.
That's why standardization matters as much as the test itself: the same distance, the same instructions, the same start and stop rules, captured the same way at every visit. The less variability the process adds, the smaller the real change a clinician can confidently see.
Trends beat single comparisons
One before-and-after pair is the weakest form of evidence. A series of consistent measurements tells a far clearer story: a steady upward trend across several sessions is more convincing than any single jump, and a sudden outlier stands out for what it is. Frequent, low-burden re-testing — including guided re-tests at home — gives clinicians more points on the line and fewer reasons to over-read one number.
Putting it into practice
- Pick the measure for the goal. Gait speed, TUG, sit-to-stand and balance tests each answer different questions.
- Standardize the protocol and document it, so every re-test is comparable.
- Interpret against published thresholds, then weigh them against the patient's own goals and presentation.
- Look at the trend, not just the latest value.
Objective, consistent capture doesn't decide what a change means — the clinician does. What it provides is a cleaner signal, so the conversation with the patient, and the note in the chart, rests on progress that's actually there.