This post is intended to explain the actual primary results of the PACE Trial (as opposed to the spun results.)
All questions, comments, feedback and corrections are very welcome.
Hi Bob
Great work on all of this. Afraid I've only just got round to printing it out and readling it all, but it's a very good summary of the PACE trial and the way it's been misrepresented. Thank you.
I'd like to make some constructive suggestions too; inevitably these take up more space than saying 'a brilliant job', which it is,
but they might help make what you've done even better. Some of the comments are about making things clearer, others in making things persuasive to those who have already accepted the 'official' PACE line.
Primary Results/Improvement rates:
I think the 'Number Needed to Treat' data is exactly the right way to present the key results of PACE, as it shows how many people have to be treated for just one patient to improve by a modest amount. Curiously, Michael Sharpe used these statistics when asked to explain the results of the PACE trial, giving a very different picture to that of the Official launch.
To simplify: I would just use NNT for those improving in
both fatigue and function as this is probably the most important measure, and not bother with separate NNTs for fatigue, function. I think it would also help if you first explain what a 'Clinically Useful Difference' is, either using the 0.5 SD definition or just saying the 'minimum useful improvement' in both fatigue and function. Then present the data.
Clinical Effectiveness /Therapeutic Effect Size:
It might help just to explain that while the 'primary' results above are categorical - i.e. what
proportion of patients reach a threshold, these relate findings relate to continuous measures i.e. how much the
average patient score improved.[/quote]
Presenting the data in a neutral way
Maybe it's just me, but I find evidence far more persuasive when it's presented in a straightforward way, without pushing the data to appear truly dramatic. The PACE results are very poor when viewed this way, which is why I'm not comfortable hyping the findings a little:
This means that approximately 87% of patients did not benefit from CBT or GET, the only treatments recommended for ME by NICE.
By contrast, 58% (physical function) and 65% (fatigue) of the SMC group achieved a clinically useful outcome.
Technically, this is more or less accurate, but it does give a slightly skewed picture, which I will try to illustrate with an example. Let's say 60% of the SMC group improved and instead of GET/CBT, the trial was of NewWonderDrug. And let's say 95% of patients on NewWonderDrug improved, a pretty decent result. The approach in quotes would present that as "65% of patients did not benefit from NewWonderDrug/Only 35% improved. By contrast 60% of the SMC group showed a clinically useful outcome". I'm not sure that's a sound way of presenting the data.
I still think the NNT figures are the fairest way of presenting the data. Sure, the PACE authors tried to present the results in a way that wasn't fair or accurate but I don't think we should follow their lead on this. The naked results are powerful enough on their own, in my view.
For similar reasons, I think describing fatigue as a 'self-reported subjective' symptom is over-egging things a bit. In contrast to physical function, there is no objective measure of fatigue, whoever is doing the experiment. I think it's appropriate to point out once that it is a self-reported measure and so liable to reporting bias, but to continually label a subjective phenomenom as subjective might irritate some people you want to influence. Again, I think the argument is strong enough without embelishment.
That's it. I think you've done a fantastic job here, and it must have been a huge amount of work. I just think if you rein it in a little in a couple of places it will be even more persuasive. 'Less is more'.