Low back pain with sciatica is one of the most common complaints for which patients seek medical advice, and the condition has considerable economic consequences in terms of healthcare resources and lost productivity. Most patients return to their normal activities within six months. For many patients, such as the one described below, a lengthy spell off work can have serious repercussions, and a rapid return to work is imperative.
A man of 27 presented with low back pain of acute onset that was complicated by pain radiating down his right leg. The problem had coincided with a recent spell of gardening and had worsened progressively over the next few days, ultimately developing into numbness and tingling of the right foot. Although analgesics had controlled the pain, his foot had become noticeably weaker. At this stage he had consulted his general practitioner.
He was found to be fit and healthy; he had no history of similar complaints. Physical examination showed that he had reduced forward flexion of the lumbar spine; straight leg raising was limited to 45 [degrees] on the right hand side. A sciatic stretch test on his right leg had positive results, and responses to light touch and pin prick were reduced on the lateral side and over the dorsum of the right foot. Dorsiflexion of the right foot showed slight weakness (MRC (Medical Research Council) grade 4), and the right ankle jerk was reduced.
His general practitioner prescribed anti-inflammatory drugs and bed rest for one week, but as this brought no improvement, a further two weeks' rest and stronger pain relief were recommended. During this period, and through his own initiative, he had a magnetic resonance imaging scan performed privately. This showed classic disc herniation on the right hand side at L5-S1.
Two weeks later he presented at my clinic (AS) with radicular pain which was aggravated by standing or sitting for any length of time. Physical examination confirmed the previous findings and did not show any warning signs or indications of progressive involvement. It was also apparent that he had a demanding work schedule and that any prolonged absence could have serious employment consequences. He wanted treatment that would allow him to return to normal activities as soon as possible.
Fortunately, the prognosis for patients with low back pain with radicular symptoms is good: 95% of patients return to normal activities within six months. Although this information reassured the patient, it was clear that a speedy recovery was his chief aim.
My usual approach to patients with radicular pain is to encourage early mobilisation and thereafter to use my clinical judgment to select an appropriate treatment for each patient's circumstances. Because of the patient's specific requirement to return to work as soon as possible, I decided to use this opportunity to reassess my usual approach and evaluate evidence for the possible treatment options.
Formulating the question
A general overview of lumbar disc prolapse showed that the eventual outcome would be the same regardless of intervention. Although some treatments offered a better short term prognosis in terms of speed of recovery, other factors such as costs and potential risks would need to be taken into account. I usually consider several further management options in treating acute radicular back pain: epidural corticosteroid injection, chemonucleolysis, and discectomy. The ultimate decision in this case would still require an individualised approach--that is, determining which option would be most suitable for this particular patient under these particular circumstance. Accordingly, the clinical question that we posed was, which of these three treatments would offer the patient the quickest return to recovery with the fewest potential adverse effects?
Because a conservative approach had not proved successful, we restricted our research to invasive treatments. We chose to search Medline (via Ovid) because of familiarity and ease of access, and because it provides a comprehensive source of current information.
Our previous experience with Medline had involved searching through reams of irrelevant, but often distractingly interesting, articles. This had certainly taught us the value of using a specific and focused search strategy. Because of time constraints and to reduce any temptation to browse, we designed our preliminary search strategy away from the computer, using the "shopping list" rationale.
We began our search using the MeSH heading sciatica as the parent term, "exploding" it to ensure that other subterms in the corresponding tree, such as intervertebral disc displacement, disc herniation, and prolapsed disk, would be included. Because of our particular interest in clinical management, we selected the subheadings of therapy, drug therapy, rehabilitation, and surgery. Further limitation to human studies and English language resulted in 84 hits for this search.
A combination of the heading sciatica and the text word "epidural corticosteroid" gave 12 possibilities. These articles were further limited to meta-analyses, reviews, and randomised controlled trials. This resulted in two relevant articles.[3 4]
The MeSH heading and text word "intervertebral disc chemonucleolysis" was searched, using the previous subheadings; this retrieved four potential articles. These four were then combined with "sciatica" using the Boolean operand "and;' which effectively limited any articles retrieved to those containing both keywords or their synonyms. This focusing strategy resulted in one relevant large scale review.
Articles on surgical decompression were found by "exploding" the parent term discectomy, combining this with sciatica, and limiting findings to relevant publication types. Of six studies retrieved, one was directly relevant. The entire search, with retrieval of articles, took 20 minutes.
Epidural corticosteroid injection--Watts and Silagy performed a meta-analysis of the efficiency of epidural injections, using 11 suitable, good quality trials involving 907 patients. They reported a definite beneficial outcome (pain relief [is greater than] 75%) in the short term ([is less than] 60 days) and long term (up to 12 months) in comparison with placebo. The odds ratio for improvement in pain was 2.79 (95% confidence interval 1.92 to 4.06) in the short term and 1.87 (1.31 to 2.68) in the long term. Adverse effects were few ([is less than] 2.5%) and transient (dural tap, headache, increased pain). Details of a reanalysis of this study are given on the BMJ website. Carette et al, in a randomised controlled trial involving 158 patients, also reported a short term improvement in the leg pain and sensory deficits caused by a herniated lumbar disc, but this form of treatment did not reduce the need for surgery in the future. They reported very few adverse effects.
Chemonucleolysis--Nordby et al reviewed studies of chemonucleolysis with chymopapain. These included 135 000 patients studied between 1982 and 1991, and showed that long term improvement in sciatic pain was maintained. Adverse effects of chemonucleolysis were infrequent ([is less than] 0.1%) but may be more serious (that is, anaphylaxis, infections, bleeding problems, and neurological deficits) than those associated with epidural injections.
Discectomy--McCulloch reviewed studies of macrodiscectomy and microdiscectomy for lumbar disc prolapse. These showed high success rates (80-96%) and indicated that short term success was particularly good when clinical findings and radiological findings agreed. However, the long term results were only slightly better than those achieved with conservative non-surgical management. There was no important difference between macrodiscectomy and microdiscectomy, and although adverse events were few, these were more serious than those seen with epidural injections.
Applying the evidence
The evidence obtained was discussed with the patient. He had the choice between an epidural injection, which might produce only short term relief and not obviate the need for future intervention, and a more interventional procedure with a higher success rate but a greater risk of adverse events (albeit small). He opted for lumbar epidural, which was performed. He noticed a considerable improvement within a week and was able to continue with his demanding work schedule. He was followed up intermittently and was given general advice and care for his lumbar spine together with a programme of appropriate spinal exercises. When last reviewed (approximately a year after the epidural) he had no residual neurological deficit. We feel that the epidural worked for him, although sceptics might say that this was merely natural resolution.
Conservative management of sciatica may fail to bring relief, and a more invasive treatment is required
Evidence shows that an epidural injection of corticosteroids produces short term relief; adverse effects are few and not usually serious
Chemonucleolysis or discectomy have higher success rates but a greater risk of more serious adverse events
Long term results after surgery are only slightly better than non-surgical intervention
Patients require an individualised approach using best evidence and the application of clinical art and expertise
Pertinent features of the case
* Fit, young patient
* Lumbar disc prolapse L5-S1
* Minimal neurological signs
* Non-progressive condition
* Main symptom of pain
* No response to conservative management
* Absence from work could have serious consequences
We thank Professor David Sackett and an anonymous reviewer for their helpful comments.
Competing interests:None declared.
 Saal JA. Natural history and non-operative treatment of lumbar disc herniation. Spine 1996;21 (suppl 42):2-9S.
 Sackett DL, Richardson WS, Rosenberg W, Haynes RB. Evidence-based medicine: how to practice and teach EBM. London: Churchill Livingstone, 1997.
 Watts RW, Silagy CA. A meta-analysis on the efficacy of epidural corticosteroids in the treatment of sciatica. Anaesth Intens Care 1995;23:564-9.
 Carette S, Le Claire R, Marcoux S, Morin, F. Epidural corticosteroid injections for sciatica due to herniated nucleus pulposus. N Engl J Med 1997;336:1634-40.
 Nordby EJ, Fraser RD, Javid MJ. Chemonucleolysis. Spine 1996:21:1102-5.
 McCulloch JA. Focus issue on lumbar disc herniation: macro- and micro discectomy. Spine 1996;2 l(suppl 42):45-56S.
(Accepted 9 February 1999)
Department of Rhuematology, Leicester Royal Infirmary NHS Trust, Leicester LE1 5WW Ash Samanta consultant
Jo Beardsley research assistant
Correspondence to: A Samanta
COPYRIGHT 1999 British Medical Association
COPYRIGHT 2000 Gale Group