Back Pain - Treatment Success or Failure?
Dr J C D Wells, Liverpool, UK
I guess I am being asked to do this talk because of my Chairing a very well-attended and fun workshop in Vienna, at the IASP meeting in August 1999. I was the Chair, or referee, in a big punch-up (sorry, debate) between Gordon Waddell, seconded by Chris Main, and Nik Bogduk. I guess it's a lot cheaper to get me over here than Nik Bogduk. So I am going to consider these two authors (1,2) and their opinions on back pain, what causes it and what treatments are effective, and I will have to use some of the refereeing techniques discussed and described by Andrew Moore (3) and his colleagues, and indeed I will go through some of their considerations of evidence-based treatment for back pain. Also, I knew that Marshall Devor would be speaking about some of these theories and Amanda Williams has just described the inherent difficulties in outcome assessment.
When we think about pain in cancer versus pain in the back, there seem to be more differences than similarities. If we think about outcome of treatment, we would expect be able to treat successfully most patients with cancer pain to alleviate their pain and suffering, to minimise side-effects, to reduce disability and distress and to keep things controlled until the cancer is successfully treated or the disease gets the upper hand and the patient succumbs. In chronic back pain we know we face an epidemic. Well, not an epidemic of back pain itself, because that is at the same level it has always been. What we have is an epidemic of back pain disability. When we look at the data that's available on disability from back pain, it looks as if we must be doing things wrong. How can we be said to be treating pain successfully, when disability is escalating? Here are figures going across from the start of the National Health Service in the UK, when treatment became free, through our explorations of basic science and research, our discovery of the disc, better diagnostic facilities, better therapeutic facilities and so on, culminating in such a terrifying rise in disability that, if this continues, the whole of the population of the UK will be off work by the year 2017.
It is not a unique British phenomenon - the same is happening in the USA, in Europe and in particular Scandinavia and of course Sweden.
Of course, disability isn't a medical issue, it's a social one. The reason that the figures in the UK escalate so alarmingly is not because back pain is badly treated but because the Government has made it easier to claim benefit for back pain and because society no longer feels that people over 50 should continue working in heavy manual jobs. However, no provisions are made for retirement, the only provision is for you to go off sick with your back. Likewise policemen, firemen and nurses, all in high-pressure, satisfying and sometimes dangerous jobs, if they are rank and file, can go off with disability from back pain whilst no efforts are made to improve the quality of their job and issues around this.
In fact, one of the fascinating things that is happening in Britain at the moment is that, alarmed by the increasing amount of disability and disease in nurses with back pain, and alarmed by frequent litigation, a great deal of work has gone into manual handling. That is, there is a hypothesis that nurses hurt their backs because they lift patients incorrectly. If they have different methods of lifting patients, ie, with hoists, extra staff and so on, this back pain and this disability will be avoided. Of course, Waddell, in his book, covers very carefully this very point and makes the issue that occupational factors are not often of any significance in the production of back pain (4).
Now I am going to stop talking about back pain for a bit and talk about my suitcase. Here is a picture of my suitcase at Manchester Airport, before coming here. You can see it is carefully labelled. Let's have a look at this label and see if it is useful. Oh, it says "suitcase". Well, that's a bit like the time I got totally lost in London, and I was driving around, even though I had a map. I stopped, opened the window and asked a passer-by where I was. As luck would have it, he was a politician. He answered carefully in measured terms "Why, you are in a car". That's a perfect political answer - short, truthful and not in the slightest bit helpful. Labelling my suitcase as a suitcase is not in the slightest bit helpful. Labelling low back pain as "low back pain" is not in the slightest bit helpful either.
I can get clever, so I am going to change this label and put something else on. Now the label says "Dr J C D Wells", and that is actually more helpful, because if my suitcase gets lost now it might actually end up with me. However, it is not very helpful to the baggage handler.
Finally, the helpful girl at the desk managed to put a more appropriate label on. This says "COP" which, if you understand the abbreviation, means "Copenhagen Airport" which is where my suitcase is going to. Now this is a meaningful label, because this tells the baggage handler which plane to put the suitcase on and so at the other end there is a very good chance that my suitcase will arrive, even though this particular suitcase has a bad track record of getting lost, but in fact it actually made it on this flight. The "Dr Wells" label is the spurious label that is put on back pain, such as "facet joint disease", "degeneration" and so on, based on the therapist's own cognitions and prejudices. The label put on by the girl at the desk is a more evidence-based label, assessed from my ticket. Now the boarding card I have issued and the suitcase label are the same, so there is a chance of success. Otherwise we are only going to have failure.
Now here we have a dichotomy between Waddell and Bogduk. Waddell states that in 85 per cent of patients with chronic back pain the cause is not known. That is not to say there is no cause, but merely that we cannot identify the specific problem. He is going on the history. In that case, this is completely true. Bogduk takes things a step further and says that provocative testing, with diagnostic facet joint injections using different types of local anaesthetic can identify specific structural causes beyond the history, beyond examination (which neither author feels is particularly useful in the bulk of simple back pain cases) and in X-ray investigations. Bogduk also feels that the sacro-iliac joint has a part to play in some pain sufferers and specific disc problems over and above disc prolapses. Here of course there is a division between the exercise therapists and the needle jockeys. All the needle jockeys say "you have back pain and I know the cause" whereas the other guys are saying "I don't know the cause but you are unfit and if you get fit the pain will go away". Therein lies the question - who is right? Well, nobody knows yet. However, both authors and everybody else would agree that unless we know the cause we cannot apply treatment. If we do apply treatment and it is not for that particular cause it is not going to work. We can consider trying to treat all sore throats with antifungals; this will work wonderfully for the people with candidiasis but not for anyone else, and overall, our evidence-based medical colleagues will conclude that antifungals have no part to play in the management of sore throat. Well, it doesn't if it is just given for a symptom, but it does if we make the correct diagnosis.
No-one who manages cancer pain would dream of just sticking in treatment the minute the patient walked in and said "I've got cancer and I've got pain". They would make a diagnosis first. Is it neuropathic, visceral or soft tissue? Is there infection? This is all well-known and well established, but we don't do this with back pain. No wonder we continue to fail in our efforts to diagnose to treat successfully, when we continue to diagnose incorrectly. What about treatments? First of all, Chapter 15 in Waddell's book is entitled "Rest or stay Active?" (5). I don't think anyone would argue against this at this conference. Some Orthopaedic Surgeons still argue against this in the UK; they still claim that bed rest, traction and corsets are effective, and some of them probably believe that the Earth is flat. However, within every group of patients who would be helped by activity there will be one or two who will be unable to be active in spite of our lack of knowledge about the cause of their pain, and I disagree with those who state that everybody has to get up after 2 days and everybody has to go back to work after 6 weeks. This is just trying to fit square pegs into round holes. Once again, if we don't know the answers, if we don't know the cause, it is cruel to pick out some people who simply can't move and make them get up.
Waddell reviews the scientific evidence for the treatment of acute back pain. He discusses the Quebec Task Force (6), the US Clinical Guidelines (AHCPR 1994) (7) and the CSAG Report of 1994 (8,9). He gives 1 to 3 stars for each treatment.
Treatments are as follows:-
Non-steroidals. These get 3 stars. Different NSAIDs are said to be equally effective. They do, however, have serious adverse effects. They are less effective for the reduction of nerve root pain. No-one would disagree with this and the advent of the new non-steroidals, the COX2 inhibitors and in particular Rofecoxib and Celecoxib, I think will change practice. In a survey of my patients with chronic back pain, I have found that 48 per cent of 400 consecutive patients were unable to take non-steroidals because of gastric side-effects. That is to say, in the early days, when they had acute pain, it was not that the NSAIDs didn't help but that they couldn't tolerate them. Is this a factor in chronicity?
One of the fascinating things about the transfer of acute to chronic pain is a recurring amount; this is not evidence-based but it is food for thought. If we look at acute shingles sufferers, it is generally agreed that about 15 per cent will go on to develop post-herpetic neuralgia. This is neuropathic pain. If we look at people who sustain whiplash injuries and have neck pain, it is generally accepted (10) that 15 per cent go on to get chronic pain. This isn't neuropathic, or is it? Marshall Devor might argue that it is. These are nice, concise syndromes, good data is available but what about people with acute back pain going on to chronicity? Data is poor, but it seems to be about 15 per cent. If 6 per cent of the population of Britain have chronic back pain and 90 per cent have had back pain at some time in their lives, that seems to be about it. A simple episode of back pain, whether or not caused by injury, has about a 1 in 6 chance of becoming chronic. Why do most people get better and why do some continue? Work is ongoing, but no evidence exists. It may well be that one of the reasons for chronicity is a lack of pre-emptive analgesia from NSAIDs in this sensitive group. Are we going to get less back pain because of the COX2 inhibitors?
What about Paracetamol/Paracetamol combination drugs? The level of evidence is 2-star. What about strong opioids? Waddell suggests that there is no place for these, and that they cause side-effects. Certainly, for simple back pain which can be managed in other ways, that might be the case. However, there is evidence, albeit weak evidence, that opioids for nociceptive back pain can be helpful and reduce pain whilst activity is recommenced and this can be a useful therapeutic measure and also that they can be useful in chronic, non-malignant, nociceptive syndromes of back pain.
What about anticonvulsants and antidepressants? No comment here from Waddell; McQuay and Moore are only able to find trials for antidepressants in neuropathic pain and we don't think that back pain is neuropathic. It might be if leg pain is included, and Chan Gunn and Marshall Devor both believe that some of it might represent neuropathic pain. There is still no evidence but I am sure that a lot of patients with chronic back pain get given antidepressants and certainly those with leg pain do get anticonvulsants. The same authors also mention the use of Capsaicin cream for osteoarthritis and point out the number needed to treat for a therapeutic benefit over placebo is 3.3, which is significant. So certainly this might be a therapy for patients with osteoarthritic change of the joints of the lumbar spine and sacro-iliac region.
Bed rest. Bed rest for 2 to 7 days is worse than placebo or ordinary activity (Waddell, 3 stars). Prolonged bed rest may lead to disability and difficulty in rehabilitation.
Advice on staying active. Waddell lists this as 3 stars.
Manipulation therapy. Again, Waddell gives this 3 stars within the first 6 weeks and 2 stars for more than this. Back exercises get 3 stars in acute pain. Physical agents including ice, heat and short-wave diathermy get 2 stars. Traction not working gets 3 stars! TENS is inconclusive. Trigger point injections get only 1 star and evidence suggests that efficacy is equivocal. Acupuncture also only gets 1 star. Epidural steroids get 2 stars in acute low back pain, with sciatica. They are said to be of limited efficacy in acute low back pain without radiculopathy. Facet joint injections only get 1 star, even in acute pain! That is, in spite of the published evidence.
Waddell states that contra-indicated treatments include narcotics for more than 2 weeks, benzodiazepines, colchicine, systemic steroids, bed rest with traction and in particular manipulation under general anaesthesia and plaster jackets.
What about treatment for chronic back pain, which is what I am supposed to be talking about? Well, the first thing is "prevention is better than cure", which is why we have to talk about acute back pain and the prevention of chronicity. Waddell points out that however good acute care is, there will always be failures, and even a small failure rate will produce a large number of chronic pain sufferers with low back pain, because of the scale of problem. There is also a huge pool of existing patients with chronic low back pain who will continue to need care. How's the evidence here?
Manipulation (15) appears to be more effective than placebo (3 stars).
Epidural steroids. The argument is that these work for those patients with root pain, but not for those with chronic low back pain. Waddell doesn't star this one way or another. McQuay criticises Koes' systematic review (16) and feels that the metanalysis by Watts and Silagy (17) is an important step forward in showing that epidural corticosteroids have an analgesic effect on sciatica compared with control. Number needed to treat for 12 weeks is 13 for 50 per cent relief which, although it looks disappointing at first, is significant and in the short term the number needed to treat is 7. Again, one has to accept that in many of these patients who had the treatment, other confounding factors might have been operating which would prevent the treatment from working in any case.
What about TENS? Most authorities agree that there is no evidence that TENS has any long-term effect in low back pain. However, the Oxford group review spinal cord stimulation, quoting Turner's trial (18) and the need for randomised trials with common design.
Although Chan Gunn claims acupuncture cures everything, and this premise has been taken up by Raj Munglani in Cambridge, most authorities believe that studies are of poor or mediocre quality and that the efficacy of acupuncture is still in doubt until good quality trials take place.
EMG biofeedback appears to be ineffective.
Exercise therapy. Many studies are of poor quality. Results are patchy but intensive exercise is possibly somewhat better at 3 to 6 months, with little difference existing at 12 months. Perhaps fear avoidance was not addressed properly in these programmes.
Pain Management Programmes. Flor et al (19) reviewed 65 reports of multi-disciplinary treatments for chronic low back pain. They concluded that these programmes can improve pain, mood and quality of life. However, these studies are criticised both by Bogduk and by Waddell. Van Tulder's study (20) is much more critical and comments that 8 of 11 trials show positive results. However, although the results are statistically significant the effect is weak. The approaches are theoretically good and the published results are encouraging, but certainly not conclusive. Thus it is perfectly reasonable to continue to try to improve our diagnostic acumen, the assessment of the factors concerned and their appropriate treatment.
Facet joint injections. Waddell only mentions this in order to dismiss these, and states that they are not effective. However, good clinical trials exist - Gallagher and Wedley, and the work of Bogduk and others (21). Bogduk of course calls these "the zygapophyseal joints" and spends a great deal of time in his book discussing them, their importance and their influence on back pain. He is in no doubt and he feels he has the evidence to prove, that these are a significant factor in low back pain and that treatment is appropriate and effective. I believe that this is a weight of evidence that cannot be totally ignored, that these joints do have a part to play in both acute and chronic back pain and that they have to be assessed carefully by appropriately trained specialists and treated on their merits.
Bogduk quotes studies by Schwarzer (22,23) (and himself) and points out that in injured workers the prevalence of facet joint disease is approximately 15 per cent and in an Australian population of elderly patients in a Rheumatology practice, the prevalence is 40 per cent. He points out it cannot be diagnosed clinically, that controlled diagnostic blocks are the only means available for establishing a diagnosis and that treatment is effective in correctly diagnosed patients. He also points out that there is a great deal of scientific evidence that the sacro-iliac joint and the disc can be a significant source of pain, even though these are difficult to diagnose. His own personal conviction is against a musculo-ligamentous aetiology in chronic back pain, although he admits that these structures can both produce pain when stimulated by trauma or experimentally.
So where do we go from here? I don't think there is any doubt that we should manage acute low back pain in the way described in the CSAG report and in The Back Pain Revolution book. I have been doing this for some time and I am sure many of you have. However, within those acute pain sufferers I think we should look carefully for those with evidence of facetal joint and other structural problems. We should accept that some patients are genuinely disabled, even though we cannot make a diagnosis. The general rule that exercise is good can be given, but cannot be applied nihilistically to every sufferer.
Society needs to decide what it wants to do with disability - personally I don't think it is a medical issue.
With regard to chronic back pain, I think it is wrong just to say that we can't tell the cause, so we can't treat it. I think that we have to make determined efforts to assess our patients who have back pain better, and to use evidence-based techniques such as those described by Bogduk, to delineate the factors involved in the pain; we should also look at other factors and other methods in the next few years.
Prevention of chronicity is better than treatment of chronicity, but careful assessment of the chronic pain sufferer, and multi-disciplinary management, including exercise, Pain Management Programmes, the reduction of fear avoidance AND appropriate medical interventions such as medication (including strong opioids) and appropriately selected blocks for suitable patients by appropriate doctors, should all be considered. I am still concerned that enthusiasts try to lump the millions of back pain sufferers all under one model. No one model exists which can explain all back pain, no one general treatment strategy exists either. In the treatment of low back pain, medicine remains an art, even though we should try to make it as scientific an art as possible. Bogduk said in Vienna (to his hypothetical chronic back pain sufferer) "I would walk with you and talk with you. I would get inside your mind and see what were your worries and your concerns. I would try to understand what makes you tick. I would jog around the island with you and get you physically fit. I would also assess you for remediable causes of your pain, including facet joint problems". Gordon Waddell said "That's the most sense I've ever heard Bogduk talk". This isn't an easy model to use but it is the only one to use! Thank you for your attention.
11. Koes BW et al 1996b Efficacy of NSAIDs for low Back Pain: A systematic Review of randomised controlled Trials of 11 Interventions. In: Val Tulder MW, Koes BW, Bouter LM (Eds) Low Back Pain in primary Care: Effectiveness of diagnostic and therapeutic Interventions. Institute for Research in extramural Medicine, Amsterdam, p 171-190 systematic review
13. Henry D, Limm LLY, Rodriguez LAG et al 1996 Variability in Risk of Gastrointestinal Complications with individual non-steroidal anti-inflammatory Drugs: Results of a collaborative Meta-analysis. British Medical Journal 312: 1563-1566 systematic review
20. van Tulder MW et al 1996b Conservative Treatment of acute low Back Pain: A systematic meta-review of 81 randomised controlled Trials of 11 Interventions. In: van Tulder MW, Koes BW, Bouter LM (Eds) Low Back Pain in primary Care: Effectiveness of diagnostic and therapeutic Interventions. Institute for Research in extramural Medicine, Amsterdam, p 171-190 systematic review
22. Schwarzer AC, Aprill CN, Derby R, Fortin J, Kine G, Bogduk Nl 1994 Clinical Features of Patients with Pain stemming from the lumbar zygapophyseal joints. Is the lumbar Facet Syndrome a clinical Entity? Spine 19: 1132-1137
23. Schwarzer AC, Wang S, O'Driscoll D, Harrington T, Bogduk N, Laurent R 1995 Prevalence and clinical Features of lumbar zygapophyseal Joint Pain: A Study in an Australian Population with chronic low Back Pain. Ann Rheum Dis 54:100-106
van Tulder MW et al 1996 Conservative Treatment of chronic low Back Pain. A meta-analysis of 80 randomised Trials of 14 Interventions. In In: Val Tulder MW, Koes BW, Bouter LM (Eds) Low Back Pain in primary Care: Effectiveness of diagnostic and therapeutic Interventions. Institute for Research in extramural Medicine, Amsterdam, p 245-285
|Up to top of page|