Further commentary on the PACE trial: Biased methods and unreliable outcomes

D

Dolphin

Senior Member
Messages
17,568
In PACE, modest improvements observed in the CBT and GET groups (contested by reanalysis) are not mirrored by substantive changes in objective measures of walking ability on a 6-minute walking test or step test (McPhee, 2017). Adding CBT to SMC did not substantially improve function from baseline (McPhee, 2017). In addition, the PACE authors dropped plans to assess patients’ physical activity using electronic monitors (actometers) on the grounds they were too burdensome. Other measurements of physical function were not considered, such as measuring how many hours per day a participant spends upright, or in bed, or laying down (pre- and post-treatment). In addition, there is almost no change in secondary measures (employment or health care use) in CBT or GET groups (McCrone et al., 2012). Such data suggest recovery in PACE is more a design artefact than a clinical reality.
Click to expand...
Strictly speaking, the aforementioned data doesn't tell us anything about recovery. We would need specific data on the subgroup: a particular subgroup could potentially have good outcomes on secondary measures such as employment or health care use while others did worse on average.
 
B

Barry53

Senior Member
Messages
2,391
Location
UK
What is clear is that when authors publish "results" that is not really what they are doing ... the are publishing their interpretation of the results ... what else can they do in fact. No matter how rigorous their science may be, there will unavoidably be some assumptions - perhaps bias - creep in unwittingly simply by virtue of being human. Good scientists will encourage scrutiny to help identify that, even though it may be difficult for them. Which is why, when so many people's lives are at stake, the data must be openly available, so that:-
  • The authors' original interpretation of results can be critiqued.
  • Other interpretations of results can be ventured if applicable.
The other massive benefit of clearly stating at the outset that anonymised data will be made openly available, is it would clean up the whole act around clinical trials. The PACE crowd would probably have been put off in the first place, and those undertaking such trials would know not to monkey around, and just keep it clean.

And there can never be any excuse that data cannot be made openly available due to patient confidentiality arguments. If the requirement is specified into the protocol from the start, then the trial's data design can readily ensure good and safe segregation between the anonymised and non-anonymised data components.
 
Luther Blissett

Luther Blissett

Messages
724
Location
Yorkshire, England
Barry53 said:
No matter how rigorous their science may be, there will unavoidably be some assumptions - perhaps bias - creep in unwittingly simply by virtue of being human. Good scientists will encourage scrutiny to help identify that, even though it may be difficult for them. Which is why, when so many people's lives are at stake, the data must be openly available,
Click to expand...

Yes!

The one of the points of the renaissance was that doubt, about oneself and others was good. It is supposed to be one of the foundations of our society.

The citizen's job is to be rude - to pierce the comfort of professional intercourse by boorish expressions of doubt. John Ralston Saul
Click to expand...
 
slysaint

slysaint

Senior Member
Messages
2,125
Haven't looked into this too much but am I right in thinking that the MRC are not the governing body when it comes to clinical trials? I could only find guidelines from them for overseas trials, that say they do NOT cover the UK.
and their guidelines (2004) for UK trials only seem to apply to those funded by them(?)
https://www.mrc.ac.uk/publications/browse/good-research-practice-principles-and-guidelines/

But I did find this from 2011: the UK Clinical Trials regulations from CHCUK
http://www.chcuk.co.uk/pdf/2011-09-01_UK_Clinical_Trials_Regulations-(Stuart_McCully)_Low_Res.pdf

which includes quite an interesting history of Clinical Trials Law and a section on European directives and
Conditions and Principles which apply to all Clinical Trials in the UK(p19),
and GCP resource document created for NIHR.

Is there anything more up to date?
 
You must log in or register to reply here.
Back