Designing and completing a large, well-designed randomised study is hard work. It must be very depressing when the result is negative – your programme of rehabilitation did not have any measurable effect. It happens. But we should not despair, and we should learn, and profit from failure. (The studies are here and here.)
The first obvious but often overlooked positive feature is that resources will be saved for other patients and treatments. This was not a cheap intervention (but back pain is long-term and very expensive, so it was important to try it). The trial was large enough and well-designed, and gives a definite answer. Many services were and may still be offering programmes similar to this. Some people may have been considering researching it. All these activities can now stop, and the resource used elsewhere.
This study’s negative outcome has saved much money and therefore has, indirectly, benefited many patients – but not patients with low back pain.
The second potentially positive feature may also easily be overlooked. The temptation is to say “this does not work, let’s move on“. That would be to waste the resource already expended. Failure was not attributable to the design or execution. It is vital to ask, why did it have fail?
One possible reason is that the intervention lacked a sound justification. Unusually, for a rehabilitation study, this intervention has been published (here). It would seem that the intervention had a sound theoretical basis, and the evidence behind the theories is at least reasonable if not good.
A second possible reason is that there was a failure to deliver the rehabilitation planned. There is no evidence of major failure.
A third possible reason is one the researchers were obviously aware of. The rehabilitation programme gave significant attention to transferring any learning and change to the home environment. The attempt failed. Why?
One way to look at the programme is as follows:
- the person at home needs to change behaviour, attitudes, beliefs etc.
- this change can be achieved in hospital
- we then need to preserve the change while transferring it back home
- we set up a transfer process.
So the two apparent weaknesses are as follows. The hospital programme failed, or the transfer programme failed.
But there is one other possibility. In the argument above, the home environment is seen as a neutral environment. Change the person, and preserve that change, and all is solved.
In reality, the home environment (in its broadest sense) was and remains the primary cause for the back pain in the first place. Far from being a neutral environment, it is an active causal environment and any benefits will be actively reversed (not in a conscious, willed sense!).
So my hypothesis arising from this analysis is that all programmes for people with back pain that depend upon removing someone from their home will fail. There may be detectable change during the inpatient stay. Indeed there almost certainly will be detectable change, even if there is no programme, and a two-week holiday on a tropical island might be even more effective at two weeks. But six months later the situation will have reverted to its earlier state.
Instead, the effort needs to be put into two things.
The first is a close, critical and detailed analysis of why the situation exists for the particular patient, identifying probably several factors that precipitated it and/or, and more importantly, factors that maintain it. The analysis needs to ficus on factors that can be altered. These factors are likely to be social and psychological, and almost certainly will involve people and organisations outside the home.
The second is to plan, with the patient and others, (a) how the patient can be helped to manage and cope with factors that cannot be altered and (b) how the physical, social, financial and other contexts can be changed to facilitate the patient’s improvement.
On reflection, having written this, I would suggest that the original idea was still based in a biomedical model where a single treatment (albeit very complex and multifactorial) was ‘applied to’ the patient so that he or she ‘was cured’. If anything, the ‘treatment’ has to be ‘applied to’ the patient’s environment so that the patient ‘cures himself’ naturally.
In summary, failure can save many resources, benefit many patients, and, if used properly, lad to a change in the understanding of the original problem and this, possibly, to an effective intervention.