Intended for healthcare professionals

Education And Debate

Controversies in Management: The case for spinal fusion is unproved

BMJ 1996; 312 doi: (Published 06 January 1996) Cite this as: BMJ 1996;312:39
  1. James Wilson-MacDonald, consultant orthopaedic surgeona
  1. a Spinal Unit, Nuffield Orthopaedic Centre, Oxford OX3 7LD

    Back pain affects about 80% of the population at some time in their life, and it is generally agreed that most people are best treated conservatively. Most episodes of acute back pain will settle either spontaneously or with treatment, but in a few people the pain does not resolve and they may be considered for spinal fusion.

    Spinal fusion was originally developed for the treatment of tuberculosis, poliomyelitis, and scoliosis. Because it seemed successful for treating back pain arising from these conditions, it was assumed that any mechanical back pain could be successfully treated with spinal fusion. More than 20000 lumbar spinal fusions are performed annually in the United States, and there is a direct relation between the number of spinal operations performed in any one area and the number of orthopaedic and neurosurgeons.1 About eight times as many spinal surgical procedures are carried out in the United States per capita as in Britain.

    Conservative treatment

    Prospective studies have shown reasonably convincingly that time off work in patients with first episodes of acute back pain can be reduced by 30-50% by early physiotherapy.2 Chiropractic treatment has also been shown to have a similar success rate.

    Treatment for chronic back pain is not so convincing. Koes found that only four of the randomised controlled trials of exercise therapy for back pain were of an adequate standard to draw any definite conclusions, and it was impossible to decide whether exercise therapy was better than any other type of treatment, or whether one type of exercise was better than another.3 However, some studies have suggested that conservative treatment can be at least as successful as any surgical treatment. Kohles et al showed that up to 85% of patients with chronic back pain can return to work with intensive retraining methods.4 They used a method called “functional restoration” in which a team of professionals assesses and treats patients with chronic low back pain. Trunk strength and spinal range of motion were used as objective measurements of progress and success.

    Back pain is the commonest cause of loss of work among people under 45, with huge economic cost, but there is little incentive in the health service to apply resources to intensive retraining programmes for these patients.

    Spinal fusion

    Spinal fusion for back pain has never been compared with non-surgical treatment in a randomised controlled trial.5 The only non-randomised prospective trials of fusion have compared surgery with and without fusion for herniated discs. Of all the papers on spinal fusion for back pain, only 47 had adequate follow up, and although 68% of patients reported in these papers had a satisfactory outcome from fusion, the success rate varied between 16% and 95%.5 There was so much variability in the methods of treatment, selection, and assessment of patients that it was impossible to compare the results or to draw any conclusions about the appropriateness of fusion for back pain.

    Lumbar fusion is probably contraindicated in many circumstances. For example, the association of smoking with pseudarthrosis and failure of spinal fusion is well documented. A recent study of people seeking compensation for a work induced injury showed that the back pain and lifestyle of most deteriorated after spinal fusion.6 Fusion rates can be improved with new techniques such as segmental spinal stabilisation using pedicle screws,7 and there is a correlation between fusion rate and resolution of pain.5 These implants make the surgery more technically demanding and more satisfying to the surgeon, but new techniques often bring with them new complications. In a survey of over 600 surgical cases, 21.2% of patients had serious complications such as nerve root damage.8 Hospital costs alone for failed spinal surgery in the United States are estimated to be about $2bn (pounds sterling1.3bn) a year.

    Some evidence suggests that many conditions which cause back pain can be successfully treated with spinal fusion. For example, in young patients with spondylolithesis spinal decompression with simultaneous spinal fusion was reported to give more satisfactory results than decompression alone.9 Similarly spinal fusion for back pain and spondylolithesis can give good results, especially in children.10 However, even these procedures have not been subjected to randomised trials. Surgery to replace the lumbar discs or operations which change the balance of the spine with posterior ligament devices may prove more effective than spinal fusion in the long term treatment of patients with back pain.


    The case for spinal fusion for back pain has not been proved. Excellent results have been obtained with nonsurgical treatments, and adequate resources have not been applied to fully assess these methods. Spinal fusion probably has a place in the treatment of back pain in certain circumstances, but many patients selected for spinal fusion at the present time might be better treated in other ways.

    A multicentre randomised prospective clinical trial is currently being planned, and this will answer some of the questions regarding the efficacy of spinal fusion and other surgical procedures in the treatment of patients with back pain. Total joint replacement has almost replaced fusion of the hip or the knee for arthritis, and in future spinal fusion may be replaced by other more effective treatments in much the same way.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.