Gathering Evidence to Understand If Pain Treatments Are Effective and Safe: A Chat With Christopher Eccleston

Christopher Eccleston is a professor of medical psychology at the University of Bath, UK, where he directs the Centre for Pain Research. He is also coordinating editor of the Pain, Palliative and Supportive Care Cochrane Review Group, an effort to bring together evidence about treatments for chronic pain and reduce uncertainty about the effects of those treatments. RELIEF executive editor Neil Andrews spoke recently with Eccleston to learn more about what the evidence says about drug and non-drug treatments for chronic pain. Below is an edited transcript of their conversation.

What is evidence-based medicine?

Evidence-based medicine is a movement within healthcare that emphasizes the evidence for the effectiveness of a particular intervention—whether the evidence shows that the intervention helps people or harms them. There has been a revolution in the last 40 years where we understand that single studies looking at the effectiveness or safety of a treatment are not enough to get an overall picture. Evidence-based medicine is about bringing together all the evidence from a particular field to get a true picture of what the evidence is for a treatment being effective and safe.

People who aren’t familiar with this area might be surprised that “evidence-based” is a term that would be needed to describe medicine, because it suggests that some treatments don’t have data to support their use.

I agree wholeheartedly that it’s a strange phrase. But much of medicine is not based on high-quality evidence. I hope that eventually we can stop talking about evidence-based medicine, as a concern for the evidence becomes integral to our considerations. I think about this in three ways.

First, there are treatments we have always used that have never really been tested, and it’s unclear whether they are effective outside of one’s individual practice—are they only effective in my hands rather than everybody’s hands? We’ve started to say that maybe, in the modern age, we should ask the difficult question of whether or not the things that we’ve always done have evidence to support their use, and if they don’t, whether we can improve them.

Second, we live in a time of unprecedented progress where people are introducing new treatments and new ways of approaching old problems. We have to apply the highest possible standards of inquiry and provide the evidence, not only for effectiveness, but also for safety, to decide whether or not we should bring those treatments into common, modern clinical practice.

Third, there are treatment areas where it’s just terribly difficult to get a clear picture of what the evidence is, often because there are conditions that are so unusual such that there isn’t enough experience with the treatment, or because we haven’t brought together all the data. That’s where we need another form of inquiry where we can be critical, transparent and ask questions about what we’re doing. Evidence-based medicine is about maturity, about asking questions concerning whether what we’re doing is the right thing or not, and listening to the answers, however uncomfortable they may be.

Do you have a message for patients who might be disturbed to learn that it is still unclear whether certain treatments that physicians may already use have a solid evidence base to support them?

In my experience, generally patients are the primary experts in their own treatment and physical condition. That said, I think it’s important to be honest, transparent and open with each other so that when we’re trying a particular treatment we recognize that there are no guarantees. Some people are surprised by this, but hopefully they will feel relieved that when a new treatment comes along we are not blindly accepting that it’s effective or not—that we need to be critical and skeptical and only introduce treatments that are safe and that we know have a chance of helping somebody. I agree that at first glance the idea of evidence-based medicine might be disturbing and worrying, but with some thought and discussion, most of the people that I work with are reassured that we are bringing science to these questions rather than blind faith.

What is the Cochrane Collaboration?

The Cochrane Collaboration was born out of a realization that there was no easy way of bringing together data from a number of different sources. It’s an international collaboration that was developed not only for pain and palliative care but across all aspects of medicine, where people try to pull together all of the evidence in the worldwide literature—not just studies in the UK, for example, or the US, or Japan, but pulling all studies together and trying to aggregate this base to answer a question.

Cochrane also tries to understand the influence of bias and the methodological quality of studies. These are separate, but related factors. Science is a human production, and like all human productions, it’s influenced by all sorts of biases, received wisdom and practices. And the decisions that scientists make in the way that they do research can have a far-reaching effect on the final outcome of a study. This again points to the benefit of not just looking at one study in isolation, but looking perhaps at 20 studies together, so that we can really begin to understand why we might get a varied picture. Based on this approach, one can make a statistical estimate for how many people are likely to benefit from any particular treatment being offered.

The principles of Cochrane are to drive at all times for accuracy, comprehensiveness, and transparency of the evidence. This includes saying clearly ‘We don’t know’ when there is no way to reduce uncertainty. Cochrane has now reached 20 years since it was started, its methods have developed over that time, and it has become a worldwide collaboration. It’s likely to improve even further with improving access to patients.

What is your role in the Cochrane Collaboration?

I am coordinating editor of the Pain, Palliative and Supportive Care Cochrane Review Group, which has quite a broad scope involving both acute and chronic pain, as well as palliative care, including end of life care. In these areas, we now have approximately 400 titles in the Cochrane Library—that’s the database of evidence—at different stages of development. My principle has always been that if somebody is selling, promoting, or seeking a particular treatment, then the Cochrane Library should have something to say about it—ambitious perhaps, but the idea is that we are a trusted resource and we are trying to help.

The library now has titles based on pharmacological treatment of postoperative pain, and we also have quite a lot of titles for chronic pain, including psychological treatments. It’s hard to summarize an entity like this with so much data that’s available, but I would encourage people to look at the Cochrane Library. There are plain language summaries as well for all of the interventions that we have.

What conclusions has evidence-based medicine reached regarding pharmacological treatments for chronic pain?

My colleagues and I recently wrote a paper in The British Medical Journal called “Expect Analgesic Failure; Pursue Analgesic Success.” The general principle of that paper was the recognition that there is no one treatment that’s effective for everybody with chronic pain. It’s worth saying a little bit more about this, because there is some skepticism about evidence-based medicine, and rightly so; the danger is that we come up with answers that say either we’re not sure about a treatment or that we can only treat a small number of people. In isolation those things can be true, in that the average effectiveness of pharmacological treatments we have for chronic pain can only help maybe, at best, three out of ten people.

That seems rather bleak, certainly if you’re a patient with chronic pain, to say that results show that only a minority of people can be helped by a treatment—that seems like it’s not a very positive message, and not what you want your healthcare professional to tell you. But the truth is that there is no such thing as someone who responds to all medicines or responds to no medicines. The biggest challenge we have in pain is to understand what works for whom and in what order. We need studies that examine when to start, switch, and stop treatments.

Let’s shift gears by looking at pain psychology, which is one of your areas of expertise. In what ways can psychology help us to understand chronic pain?

Psychology can help in at least three ways. First, we do quite a lot of work here at the University of Bath trying to understand the cognitive effects of being in pain. Attention underpins pain perception—if you aren’t attending towards a sensation, then you can’t report on its severity. We’ve been working to understand the attentional system and whether through the repeated experience of and exposure to painful sensation the attentional system becomes adjusted to become vigilant for pain, and also to understand the role of attention in analgesia [pain relief].

Second, we need to think about heterogeneity, both of the patients who are asking for help and about the treatments that we have to offer, as a positive rather than as a negative thing. What I’d like to see is a much more mature dialogue around having multiple treatments for multiple types of patients. We’ve just started to do that by looking at psychological interventions for chronic neuropathic pain [pain caused by nerve injury]. What you find is that the presentation, for example, of someone with painful diabetic neuropathy is very different from that of patients with other types of pain. We need to develop the right treatments for specific types of patients, and what’s missing at the moment is that specificity.

Third, I would like to see a bridge between pharmacological and non-pharmacological treatment. For example, people are beginning to realize that the communication around pharmacological treatments, the expectations that people have about whether a treatment is going to be effective or not, and other psychological factors all need to be better understood. I hope that the future of pharmacological and non-pharmacological treatments will be less divided.

What have we learned from the Cochrane Collaboration about psychological treatments for chronic pain?

There are treatments that have been developed, over the past 40 years, called cognitive behavioral therapies, because they focus on action in the world. Over 50 randomized controlled trials have been attempted to see whether these treatments are effective in helping people with chronic pain. My colleagues and I worked to pull together all of those trials, and asked whether psychological therapies help to reduce chronic pain, reduce disability from chronic pain, and improve mood. We also asked whether these therapies are better than other therapies, or better than doing nothing.

What we found is that these treatments can be effective, but the average effect was very small. The issue that I raised earlier about pharmacological treatments is the same issue that we have for psychological interventions. That is, it would be wrong to say that everybody will receive a very small benefit, because hiding in those averages is that there are some people who can get a major benefit from these treatments that they couldn’t get in any other way, and for some people the treatments have no effect whatsoever. Again, what we need to learn now is how to develop treatments that are likely to be effective for more specific patient populations, and move away from the idea that these are treatments that will work for everybody.

The evidence that the Cochrane Collaboration considers is evidence from clinical trials, which take place in carefully controlled environments that are different from what happens in actual clinical practice—when a patient walks into a doctor’s office. What do you think about this issue?

Yes, doing a clinical trial is different than working in clinical practice; this is a known problem. What we are interested in for the clinical trial is controlling all of the variables that might influence treatment in some way. This is because one is trying to determine if there is an independent, unique contribution of a treatment to an outcome.

In clinical practice, you don’t try to control all those other variables—instead you try to bring them in, account for and make sense of them, and then use them in some way. In fact, when you move from a clinical trial to a real clinical practice situation, there are going to be differences, because each is answering different questions.

Any final thoughts?

There are significant challenges for us in the psychology of chronic pain. I believe progress will come in three main areas. First, we need to better integrate clinical and basic human behavioral science. Second, we need to resist fashion. Enthusiasm is important but over-selling damages everyone. Third, there is a need for greater collaboration with non-psychologists. We need a return to multidisciplinary efforts to understand the multidimensional problem of chronic pain. Working in teams requires scientists who are tolerant of their own ignorance, and who are willing to work with and learn from each other. The lesson of Cochrane is that science requires the humility to recognize that we only move ahead by working together.

Home page image credit: agorohov/123RF Stock Photo.