Let’s agree that the time-honored and most straightforward way to interpret an indeterminate test result is to simply pronounce that it confirms what you thought already. The more “messy” and imprecise a test — say, the 6-minute walk distance — the better suited it is for this purpose.
If you want to get more reality-based than that, Dolmage et al wrote this paper about 6MWD for you.
Six minute walk test has a large natural variability. Day-to-day changes can exceed 70 meters — which exceeds the expected benefit of 30-50 m after pulmonary rehabilitation, the original proposed minimally clinically significant difference of 54 m, and the proposed MCID in idiopathic pulmonary fibrosis of ~24-45 m. So, with all the noise, how can you tell if the change in 6MWD you observe is signal?
Unfortunately, you have to re-learn statistics, which the authors make palatable in this readable primer. Their central point boils down to: don’t apply parameters like standard deviation and confidence intervals derived from large samples to apply to your individual patient — they simply don’t. What’s needed to determine a test’s repeatability in an individual is the standard deviation on the group’s Bland-Altman plots. No one does this for us typically (or teaches us the importance of it), which is why we often flail when interpreting tests.
The punchlines are:
- The 6MWD’s repeatability is +/- 82 m by several such analyses.
- Don’t use the best value achieved by the patient — the average of repeated efforts is closer to truth.
- The more serial testing you get, the more likely truth will emerge.
- Base your judgment on all the available information, not just the 6MWD, and rely on the observed benefit in randomized trials as a metric for treatment success.
Dolmage TE et al. Has My Patient Responded? Interpreting Clinical Measurements Such As the 6-Minute-Walk Test. Am J Resp Crit Care Med 2011;184:642-646.