Given that the inferior limit of the confidence interval
around our effect size (2e99 m) lies below the limit of the
confidence interval around the estimate of the MCID for the
6-min walking test (CI: 30e42 m),13 the clinical significance
of our result is low. This finding parallels those of Boxall
et al. (CI: 8e92 m) who also included the 6MWD as an
outcome measure of a self-management education program
with supervised exercise. Their shorter follow-up period
(i.e. 3 months) renders comparisons with our trial difficult.
With a comparable 12-month follow-up, our results on
exercise tolerance and dyspnea do contrast those of two
larger randomized study from Bourbeau et al. (CI: 44 to
26 m) and Monninkhof et al. (CI: 47 to 1 m), where the
exercise program was not an obligatory component of the
intervention. It is not clear in these prior studies how many
patients really participated in the exercise program. The
standardized supervision of exercise in our intervention
(i.e. 8 sessions) may be an explanation for our better result.
Furthermore, the fact that our patients showed a similar
COPD severity (FEV1 w 1.5 L) and baseline exercise tolerance
(6MWD w 420 m) to those of Monninkhof et al.,7
makes the confounding effect of these two known factors
on the 6MWD’s change unlikely.