There are not too many conferences now in which I don’t get the question over tea and biscuits about whether or not I believe in this therapy or that. I find it an intriguing question and one I usually answer with ‘Well, not in the same way I believe in gravity or in the fundamental worth of each human’. From the look on the face of the questioner, I suspect I have missed their point. One might paraphrase the question by asking whether or not I think these therapies ‘work’. I reckon that the question often comes from a place of knowing, deep in their belly, that the therapy works for them and being confronted by an RCT or two that suggest it works a bit but not any better than this other therapy. I might say in response ‘well, let’s say it works – the question I find more interesting then, is why does it work?’
We sometimes have some pretty whacky mechanisms to explain treatments that work and I am on the record as saying that we need to be open-minded about how an apparently effective treatment might work. Well, it seems that the psychology world is dealing with the same issues. Dr John Burns has just given us his US$5 dollars worth and has argued that we have to stop doing more trials and dump the waiting list control thing and for all these mechanism-like statements we need to prove it.
To conceptualise what this means, one might state that a cognitive intervention reduces pain by reducing pain related catastrophising. Burns proposes a list of 5 criteria that must be satisfied if we are to uphold this kind of statement. Because ‘mechanism’ is terrifically difficult to measure in behavioural research (indeed the radical behaviourists clearly state that they have no interest in mechanism – only in observables), we will talk about ‘mediator’ instead.
So, to conclude that a treatment is effective because of a particular mediator, we need:
1. Change in the proposed mediator correlates with change in the outcome. If the cognitive intervention reduces pain by reducing catastrophising argument, decreased PCS score needs to correlate with decreased pain.
2. Change in the mediator precedes change in outcome. That is, decreased catastrophising occurs before the decrease in pain.
3. Early change in the mediator predicts later change in outcome but not the other way around. That is, early decrease in catastrophising predicts later decrease in pain but early decrease in pain does not predict later decrease in catastrophising.
4. Change in the mediator is specific to the treatment. That is, the control group does not show a decrease in catastrophising.
5. The relationship between change in the mediator and change in the outcome is not observed across treatments. That is, the temporal pattern of change in catastrophising and change in pain should not be observed in the control group.
Clearly, to answer these questions we will need to take serial measures – pre and post just won’t do.
As I was listening, I thought this is good stuff indeed – and most of the research in physical and psychosocial interventions for chronic pain don’t do this. Then I thought I had better think about this with my own views on my own research. Let’s start then, with Explaining Pain (EP). The theory behind explaining pain is that it decreases pain by changing the underlying schema about what pain actually is. Does this mechanism statement satisfy the Burns list?
1. Change in the proposed mediator correlates with change in the outcome. YES. Change in pain correlates with change in pain biology knowledge as measured by the pain knowledge questionnaire (PKQ).
2. Change in the mediator precedes change in outcome. YES. Change in PKQ can occur in hours but change in pain takes weeks to months. The exception to this is seen in a paper from Euro J Pain [1] in which changes in pain threshold were observed immediately. So, that experiment may not satisfy this rule but it may be because we didn’t have enough measurements over time.
3. Early change in the mediator predicts later change in outcome but not the other way around. YES. Increase in pain-related knowledge directly after a 3 hour Explain Pain predicts pain reduction 3 weeks and 6 months later.
4. Change in the mediator is specific to the treatment. YES, In the research we have done, even when we compare EP plus CBT to CBT alone, change in PKQ occurs in the EP group but not the control group.
5. The relationship between change in the mediator and change in the outcome is not observed across treatments. YES. We sometimes see a change in PKQ after a change in pain in the better control treatments (eg CBT).
So, EP satisfies the Burns test, which suggests that a change in pain-related biology knowledge contributes to the effect of EP on pain reduction.
About Lorimer Moseley
Lorimer is NHMRC Senior Research Fellow with twenty years clinical experience working with people in pain. After spending some time as a Nuffield Medical Research Fellow at Oxford University he returned to Australia in 2009 to take up an NHMRC Senior Research Fellowship at Neuroscience Research Australia (NeuRA). In 2011, he was appointed Professor of Clinical Neurosciences & the Inaugural Chair in Physiotherapy at the University of South Australia, Adelaide. He runs the Body in Mind research groups. He is the only Clinical Scientist to have knocked over a water tank tower in Outback Australia.
Link to Lorimer’s published research here. Downloadable PDFs here.
Reference
[1] Moseley GL (2004). Evidence for a direct relationship between cognitive and physical change during an education intervention in people with chronic low back pain. European journal of pain, 8 (1), 39-45 PMID: 14690673