How we currently measure back pain

Hi, I'm Rob Froud, and welcome to "How We Currently Measure Back Pain," which is a talk in the series "Current Thinking in Back Pain Management." I originally trained as a manual therapist, and went on to read epidemiology and medical statistics. I split my time between the Norwegian School of Health Sciences, in Oslo, and Warwick Medical School, in the UK.

So this is just an overview of the topics I'm going to cover in this talk. I'm gonna first talk about the history of low back pain measurements and so called 'Core Sets' of outcome measurements. And then I'm gonna talk about the properties of these instruments, what makes a good measurement instrument, what makes a bad one, and what we expect them to do. And finally, I'm going to talk about how we use these instruments to help us manage patients and populations with back pain. Now this talk focuses on how we currently measure back pain, but in understanding our approach to that, there may be a question rearing its head about whether we're actually measuring things that are important to patients. And I'll be covering that in a subsequent talk. And so I'm hoping if I hold your interest in this talk, you might go on to have a look at that one too.

So we've already heard in the series about how expensive back pain is. In 1998, it was estimated that back pain cost the UK economy 12.3 billion pounds. Somewhere around a tenth to a third of us have back pain right now, and pretty much all of us can expect to get it at some point in our life. The lifetime prevalence is as high as 84 percent. About four percent of the UK population takes time off work due to back pain. And this equates to 90 million days lost each year, and 8-12 million GP consults every year. So I think it's probably gonna be helpful if I try to sharpen the focus of the scope of this talk a little bit. It's going to be about trials, specifically, outcome measures used in trials. Outcome measure use in basic science, for example, might be a little bit different to what we're interested in trials. And secondarily, it's going to be about pragmatic trials. Now these focus on the benefit the treatment produces in routine clinical practice as opposed to efficacy trials which look at the benefit that's produced under ideal conditions. Now this is pretty easy to do in drug trials, if you imagine you have two drugs that are identical, to same color, same shape, and perhaps even the same taste, it's then easy to blind everybody in the trial, patients, doctors, to which is the active pill. And then providing that the trial was done well, the difference is attributable to the active ingredients. Now this is much harder to do in trials of complex interventions like surgery, manual therapy, or CBT, where what's given to the patient may consist of many different but inseparable components. And it's also difficult to blind people to what they're receiving, so most trials in back pain are pragmatic in nature. I think it's fair to say that pragmatic trials are also the most useful to informing policy, and managing patients in real-life situations. Now a lot of money is being spent on back pain trials to get these data. At today's prices, a full scale randomized control trial costs the thick end of 2 million pounds. So it was so much money being spent, it's important that we ensure we are measuring the right things. So if we start off in the 1970s, the focus at this time in outcome measurement is really on range of motion and mobility. And there was a notion that if you could move again, it meant be all well again, regardless of whether it still hurts. But in the 1980s, the focus started to shift towards patient-reported outcome measures, so called PROMs. And these PROMs measure latent domains, as well as things that are readily observable. So by latent domain, I mean, the domain of health that you and I would probably agree existed, but that we couldn't go out and directly measure. So something like depression, for example, where we couldn't see it to readily measure it, but we might be able to devise questions. So these would be questions about depression, so we could ask about low mood and in order to ensure we capture the full scale, we might ask about suicidal ideation. So it might seem odd to use questionnaires to measure back pain but essentially, we can't see back pain either. X-ray images and MRI images of someone with quite crippling back pain can sometimes look a lot like those of someone without any pain. Conversely, images taken from someone with no pain can show disc narrowing and bony growths, the sort of things that you might reasonably expect to be associated with back pain. In practice, the way doctors diagnose back pain is essentially just by asking the patient whether they have it. Hence this move towards questionnaires to measure and monitor back pain, which is essentially sensible so long as we're measuring the right things. So the '80s and the '90s saw many PROMs being developed, and it got to the point where there were so many, and so many trials that used different outcome measures that made it very difficult to compare outcomes between trials. And it was difficult for clinicians as well who might not have much familiarity with one particular scale, because they were so many of them. So efforts were made towards the turn of the millennium to standardize the outcome measures that were used in trials, and recommends so called 'core sets' of outcome measurements.