Hi oceanblue, 1. self-paced is irrelevant. Many of us go at our own pace and cross the limit - its what we do. When they are encouraged, expected and trained to push - what do you expect? Self paced under those conditions makes it very likely they are pushing too hard, especially on the test after their training.
2. I am not sure of the relevance of this. Why do you think its relevant? The SF-36 is not validated by studies looking at post exercise repeatability in ME. They might claim relevance of course.
3. I am going on the Pacific Labs research and commentaries for this. They cite that every other disease they have looked at - and there is a list somewhere - the results one day to the next were very similar, within I think about 7%. They do not see this in ME, the drop in capacity is sometimes 30%. I am aware, as you stated in another thread, that the Pacific Labs testing is more intense, but I think this is enough to show the 6MWT has to be revalidated to show it works on those with PEM before any results can be considered reliable. In fact the lack of intensity in the 6MWT is another count against its validity. That intensity is critical to appropriately investigate those with PEM.
The PACE authors stacked the deck when they dropped the actometers or decided not to use them in final assessment. They also knew, if they were following the literature, from about 2006, that repeatability was potentially an issue. They stacked the deck again by mostly including mild patients, which is an intrinsic flaw to using the Oxford definition, though I think it likely there were some moderate patients. Severe patients? Doubtfull. Very severe - impossible in my view, by definition - such patients could not participate.
Every single study using actometers has found either no improvement or a worsening of function - the 6MWT has to be interpreted in light of that. A Spanish study showing this came out just two months prior to the PACE study. I can push myself too, it doesn't mean I don't have payback and have to cut on my activity in other ways.
Similarly they want to imply it shows functional capacity. Yet the Belgians found that in 862 patients there were no cures and a worsening in function. This does not match the 6MWT implication. NO objective evidence matches the 6MWT findings. If they were serious they could have used gold standard exercise testing. They chose the lead standard instead. Its cheap, its easy, and its widely thought to be reliable without justification in this case.
Now, as an aside, I think if they recruited a whole lot of Oxford defined CFS patients to validated the test they would in all likelihood validate it - provided they ignored the test-retest requirement with full gas analysis etc. However if they tried it with CCC ME/CFS then I think they would fail to validate it. If they had managed to validate it for Oxford CFS patients, then at least we would know it was more or less reliable for this heterogenous cohort as a group, even if it were invalid for subgroups.
One of the things they would have to do is, with different definitions of ME and CFS, compare test-retest gold standard exercise testing versus the 6MWT. If they could pull that off, and show reliable results, then I would consider it validated.
One of the things an RCT is supposed to do, properly designed, is help guard against bias and confounding factors. However when this many confounding and potentially biasing factors are included, especially subjective measures and psychological intervention, a really large effect size would be necessary to get beyond the range of potential bias and confounding factors. They got a small effect size. They cite it as a great result purely on statistical significance - however bias and confounds can give you a very significant result and still be wrong, and the risk of this rises inversely to the effect size.
Variability of disease is another and very different factor - but it rarely matters in a large group as the average effect will still be reliable - if you don't have ME. We have variability too, like in MS. However post exertional declines are not VARIABLE they are DEPENDENT. An entirely different issue.
Bye, Alex
