Background: Sciatica is a symptom characterised by well-localised leg pain with a sharp, shooting or burning quality that radiates down the back of the leg and normally to the foot or ankle. It is often associated with numbness or altered sensation in the leg.
Objectives: To determine the clinical effectiveness and cost-effectiveness of different management strategies for sciatica.
Data sources: Major electronic databases (e.g. MEDLINE, EMBASE and NHS Economic Evaluation Database) and several internet sites including trial registries were searched up to December 2009.
Review methods: Systematic reviews were undertaken of the clinical effectiveness and cost-effectiveness of different treatment strategies for sciatica. Effectiveness data were synthesised using both conventional meta-analyses and mixed treatment comparison (MTC) methods. An economic model was then developed to estimate costs per qualityadjusted life-year gained for each treatment strategy.
Results: The searches identified 33,590 references, of which 270 studies met the inclusion criteria and 12 included a full economic evaluation. A further 42 ongoing studies and 93 publications that could not be translated were identified. The interventions were grouped into 18 treatment categories. A larger number of studies evaluated invasive interventions and non-opioids than other non-invasive interventions. The proportion of good-quality studies for each treatment category ranged from 0% to 50%. Compared with studies of less invasive interventions, studies of invasive treatments were more likely to confirm disc herniation by imaging, to limit patients included to those with acute sciatica (<3 months’ duration) and to include patients who had received previous treatment. The MTC analyses gave an indication of relative therapeutic effect. The statistically significant odds ratios of global effect compared with inactive control were as follows: disc surgery 2.8, epidural injection 3.1, chemonucleolysis 2.0 and non-opioids 2.6. Disc surgery and epidural
injections were associated with more adverse effects than the inactive control. There was iv Abstract some evidence for the effectiveness of biological agents and acupuncture. Opioid medication and activity restriction were found to be less effective than the comparator interventions and opioids were associated with more adverse effects than the inactive control. The full economic evaluations were of reasonable to good quality, but were not able to fully address our research question. Although individual studies raised a number of
important issues, it was difficult to draw meaningful conclusions across studies because of their heterogeneity. The economic model demonstrated that stepped-care approaches to patient management were likely to be cost-effective, relative to strategies that involved direct referral to disc surgery.
Limitations: The limited number of studies for some comparisons, the high level of heterogeneity (within treatment comparisons) and the potential inconsistency (between treatment comparisons) weaken the interpretation of the MTC analyses.
Conclusions: These findings provide support for the effectiveness of currently used therapies for sciatica such as non-opioid medication, epidural corticosteroid injections and disc surgery, but also for chemonucleolysis, which is no longer used in the UK NHS. These findings do not provide support for the effectiveness of opioid analgesia, which is widely used in this patient group, or activity restriction. They also suggest that less frequently used treatments, such as acupuncture, and experimental treatments, such as anti-inflammatory
biological agents, may be effective. Stepped-care approaches to treatment for patients with sciatica are cost-effective relative to direct referral for surgery. Future research should include randomised controlled trials with concurrent economic evaluation of biological agents and acupuncture compared with placebo or with currently used treatments. Development of alternative economic modelling approaches to assess relative costeffectiveness of treatment regimes, based on the above trial data, would also be beneficial.
Funding: The National Institute for Health Research Health Technology