With a background in a health profession not known for its love of science, I suppose when I came across evidence based health care I felt a strong sense of relief. Here, at last, was a way to work out what to do when faced with the task of helping people with pain. I was therefore a bit taken aback when I read a blog post on Science-Based Medicine here challenging my assumptions about the place of randomised controlled trials at the top of the evidence hierarchy (well, actually second from top – meta-analysis of RCT’s is at the top!). It took a moment or two of serious pondering to come to terms with David Gorski’s critique of EBHC, and to accept that he has a point. His point is that if RCT is at the top of the heap in evidence terms, it shouldn’t do so at the expense of basic science – so an RCT of homeopathy, for example, should never proceed because it relies upon an implausible underlying methodology (so how does water gain a ‘memory’ of a molecule after so many dilutions? – no, don’t rush to tell me!). Well, I can accept this view without too much of a stretch.
What I’m still pondering is how to translate research findings that use grouped data into my daily practice with an individual who, as we all are, is a unique individual with all his or her own characteristics. Even more challenging in New Zealand when most of the research I read is conducted in countries in the Northern hemisphere!
I came across this paper by Ostelo, Croft and others in a great series in Best Practice & Research Clinical Rheumatology on back pain. It takes a fairly traditional view of EBHC (calling it EBM just doesn’t cut it for me – there is a difference between medicine and what the rest of the great unwashed health care providers do!), but the paper does summarise some of the issues relevant in back pain especially.
Ostelo and colleagues identify two main problems with transferring research into practice: how evidence is generated, and how evidence is implemented. A good example is that of rest for low back pain. They report (and I’d forgotten this) that the negative effects of bedrest were first demonstrated in an RCT published in 1968. Yet in 1985 I worked in a hospital in which people were given bedrest, and some were still being placed into a plaster jacket to ‘rest’ the back until the pain settled. Ostelo et al., describe a study published in 2008 in which more than 25% of doctors surveyed would not advise activity or return to work for a person with back pain.
So, simply being published, and even being a beautifully conducted study does not make for behaviour change in clinicians!
Their first key point is that research does not reflect several aspects of clinical practice. The population may be quite different – where I work, people may come from a wide range of backgrounds, with co-morbidities and age and socio-economic backgrounds quite different from those in RCT. Many RCT’s don’t include people with previous surgeries for their back pain, yet they’re often the type of person who comes to the Centre in which I work. The intervention may not be carried out in exactly the same way as the research study – most research protocols prescribe in very specific terms how clinicians conduct their intervention – not so in ‘the real world’!
If you’ve ever read a Cochrane Review, you’ll note how many studies are excluded from analysis on the basis of not quite matching the selection criteria – not only this, you’ll have read the conclusion that ‘the majority of the included RCTs have a high risk of bias’ or ‘conclusions are weak’ or ‘methodology was poor’ – and as a result, the conclusions drawn are limited.
You’d think, from what I’ve written here, that EBHC, and drawing from the literature to ‘real world practice’ is pretty much a non-starter, and for many clinicians that is so true. I read so many reports written about patients in which unhelpful and inappropriate treatment modalities are repeated in the vain hope that by doing the same again and again, something new will happen!
Ostelo and colleagues then move to dissemination of EBHC – and the use of guidelines. Guidelines are one way of collating information from research and attempting to synthesise this in a way that clinicians can start to use it. Unfortunately (you knew I was going to say that), it hasn’t been especially effective. Some of the problems with clinical guidelines are that:
- clinicians just don’t agree with them (that’s not the way I learned to treat people!),
- systems don’t support a new way of working,
- they get ignored amongst the enormous wadge of information a clinician is ‘supposed’ to read,
- they can be greeted with suspicion depending on who is promulgating them (an insurer – it’s just a way to ration treatment and cut costs; a health professional group – it’s just a way to corner the treatment market for that profession…)
What to do, what to do… more questions than answers.
I’m not completely stumped, because as you’ll have seen on these pages, I regularly read the literature, wonder and worrit away, and finally come up with something that I feel reasonably confident to use when working with a real person. I work using a case formulation in collaboration with the person, testing my hypotheses about what is going on in an N of 1 approach. This means being scientific about what I do – taking measures before and after and at follow-up, trying to control for what can be controlled, and being systematic, but at the same time, taking into account what the person identifies as important.
I recognise that there are numerous confounds that get in the way of my being accurate in determining what ‘works’ and what doesn’t for the individual I’m seeing. I try to take this into account, but as I’ve mentioned before, there are those nonspecific ‘meaning responses’ that people have within an interaction that I can’t correct for. A meaning response is how a person views the process of health care – the meaning he or she takes from that clinical encounter.
I think this is an incredibly important aspect of healthcare, and I want to learn more about it. It’s more than simply developing rapport, or explaining things clearly or even the way I set goals or measure them. It’s about the sense that the person seeing me makes of our interaction – and it’s present before the person sees me, and goes on afterwards. It’s based on sociocultural aspects of the healthcare context, the person’s own beliefs and expectations, the mechanics of what I might do as well as what develops between us as we interact. It’s profound in its power and mysterious in its workings, and I want to learn more.
Ostelo, R., Croft, P., van der Weijden, T., & van Tulder, M. (2010). Challenges in using evidence to inform your clinical practice in low back pain Best Practice & Research Clinical Rheumatology, 24 (2), 281-289 DOI: 10.1016/j.berh.2009.12.006