The recommendations on what advice to provide were synthesized from national and international evidence-based guidelines [Waddell, 2004; Chou et al, 2007; Koes et al, 2007; Health Education Board for Scotland and Health and Safety Executive, 2008].

Advice to not rest in bed any longer than is necessary

• The advice that people with sciatica should rest in bed no longer than is necessary is based on expert opinion that keeping active can help people with sciatica cope with their pain, may help them return more quickly to full usual activity, and is unlikely to make their sciatica worse. Clinical trials (with methodological weaknesses) have failed to provide evidence that advice to rest in bed is more effective than advice to stay active, or that bed rest is more effective than physiotherapy.

Advice to keep as active as possible and exercise regularly

• The advice that people with sciatica should keep as active as possible and exercise regularly is based on expert opinion, because a randomized controlled trial (RCT) found no evidence of benefit from adding exercise to conventional treatment or adding exercise to massage/manipulation. RCTs found no statistically significant difference between exercise and traction, or between exercise and manipulation. One small RCT found that people who took exercise were more likely to be satisfied than people who received physical therapy without exercise.

These recommendations reflect British, European, and US evidence-based guidelines on the management of sciatica and low back pain [van Tulder et al, 2004; Chou et al, 2007; Koes et al, 2007; Health Education Board for Scotland and Health and Safety Executive, 2008; WeMeReC, 2008; National Collaborating Centre for Primary Care, 2009].

Analgesia

• There is no good evidence from clinical trials to guide the use of analgesics to relieve pain and disability, so the recommendations about analgesics are extrapolated from guidelines on low back pain [van Tulder et al, 2004; Chou et al, 2007; Koes et al, 2007; Health Education Board for Scotland and Health and Safety Executive, 2008; WeMeReC, 2008; National Collaborating Centre for Primary Care, 2009] and other knowledge about the efficacy and adverse effects of analgesics [BNF 57, 2009].
• NSAIDs and paracetamol
• The risk of serious adverse effects is higher with NSAIDs than with paracetamol, which is therefore the preferred first-line analgesic.
• Coxibs are currently not licensed to treat people with sciatica, but NICE recognized that practitioners might offer these to people with chronic low back pain who are at risk of gastrointestinal effects, and referred to the NICE guideline on Osteoarthritis for evidence on minimizing the risk of adverse effects, including the use of PPIs for gastroprotection [National Collaborating Centre for Chronic Conditions, 2008]. CKS has therefore also recommended them for the treatment of sciatica.
• The evidence on assessing and managing the adverse effects of NSAIDs is reviewed in the CKS topics on Dyspepsia - unidentified cause and NSAIDs - prescribing issues.
• A proton pump inhibitor (PPI) is recommended for people at increased risk of adverse gastrointestinal effects from NSAIDs.
• This recommendation is adapted from the NICE guidelines on osteoarthritis [National Collaborating Centre for Chronic Conditions, 2008]. Because the NICE guidelines on Osteoarthritis were developed to apply to people from the age of 45 years, the low back pain guideline development group clarified the wording of this recommendation to include the age threshold of 45 years for routinely co-prescribing a PPI with standard NSAIDs and coxibs.
• The evidence on the use of PPIs to prevent the gastrointestinal adverse effects of NSAIDs is also discussed in the CKS topic on Dyspepsia - unidentified cause.
• Choice of NSAID in people at risk of cardiovascular adverse effects: diclofenac 150 mg daily has a similar level of thrombotic risk to etoricoxib and possibly other coxibs. The available data do not suggest an increase in thrombotic risk with naproxen or with lower doses of ibuprofen (up to 1200 mg daily) [CHM, 2006; MHRA, 2007; MHRA, 2009]. For further information, see the section on Supporting evidence in the CKS topic on NSAIDs - prescribing issues.

Trial of amitriptyline and pregabalin (or gabapentin)

• The recommendation to offer a trial course of amitriptyline or pregabalin (or gabapentin if there is a local decision to prefer gabapentin over pregabalin) is based on guidance issued by the National Institute for Health and Clinical Excellence on drug treatment of neuropathic pain in adults [NICE, 2010]. For further information, see the CKS topic on Neuropathic pain - drug treatment.
• Having reviewed the evidence for a number of neuropathic conditions (including radiculopathies and radicular pain), the NICE guidance development group (GDG) treated the term 'neuropathic pain' as a blanket condition regardless of the underlying cause; the GDG considered this to be helpful and practical for non-specialist healthcare professionals and patients. However, condition-specific recommendations were made if robust evidence on clinical efficacy and cost-effectiveness existed (as in the case of painful diabetic neuropathy), or where the evidence was clearly uncertain and insufficient to alter current clinical practice (as in the case of trigeminal neuralgia). The GDG acknowledged that evidence for treating a particular neuropathic pain condition with a particular aetiology is often extrapolated to other neuropathic pain conditions with other aetiologies, although there is little evidence to support the validity of this [NICE, 2010].

Referral for physiotherapy

• Referral for physiotherapy is recommended for people who are recovering slowly because there is consistent evidence from three RCTs that physiotherapy may provide statistically significant and clinically important benefits for people with sciatica for at least a year. A Dutch economic analysis of one of the RCTs found that physiotherapy was not cost effective, but it is not clear if the conclusions are generalizable to the UK.

Referral for epidural injection of corticosteroids

• Referral for assessment for epidural injection of corticosteroid is recommended because this is performed by some specialists. However, a systematic review found mixed evidence, with some positive results and some negative results, from nine studies that compared epidural or extradural corticosteroid injection with placebo injection [Luijsterburg et al, 2007a]. Meta-analysis was not performed, because the studies were too heterogeneous.

Referral for specialist assessment if a strong opioid is necessary

• If the use of a strong opioid is becoming chronic, or if doses are escalating, obtaining specialist advice is recommended because of the risk of adverse effects such as dependence.

Benzodiazepine for muscle spasm

• Benzodiazepines are recommended for people with muscle spasm because:
• Although there is no evidence from clinical trials in people with sciatica, there is evidence that muscle relaxants relieve acute low back pain. For more information, see the Supporting evidence section on Muscle relaxants in the CKS topic on Back pain - low (without radiculopathy).
• In the UK, benzodiazepines are more commonly used for this indication than other muscle relaxants. Diazepam is recommended because it is longer acting and less likely to result in withdrawal symptoms than other benzodiazepines, and is the only benzodiazepine that is licensed for the treatment of muscle spasm.
• A short course (up to 7 days) is recommended because the risk of adverse effects is high and the danger of habituation is great [van Tulder et al, 2004; Chou et al, 2007].

Treatments not recommended

• Spinal manipulation
• There is insufficient evidence to recommend spinal manipulation for sciatica. One RCT found that spinal manipulation may be more effective than sham manipulation at relieving pain, and that the difference in effect could be maintained for 6 months. Several other RCTs provide weak or inconsistent evidence on the effectiveness of spinal manipulation compared with infrared heat, traction, exercise, and wearing a corset. All the RCTs had methodological weaknesses.
• Acupuncture
• There is insufficient evidence to recommend acupuncture for sciatica. Single small RCTs provide very weak evidence of possible small clinical benefits of acupuncture compared with NSAIDs and sham acupuncture.
• Massage
• There is insufficient evidence to recommend massage for sciatica. One small RCT provides limited evidence that massage/manipulation (with or without exercise) may be more effective than traction.
• Traction
• There is insufficient evidence to recommend traction for sciatica. Three RCTs found that traction is no more effective than sham traction or minimal treatment for sciatica. Five RCTs provide no good evidence that traction is more effective than other conservative treatments for sciatica. The methodology of only one study was considered to be of high quality.

These recommendations are synthesized from national guidelines and are largely based on expert opinion [NICE, 2001; Health Education Board for Scotland and Health and Safety Executive, 2008].

Referral for surgical assessment and opinion on discectomy

• One large, well-conducted, and fully reported randomized controlled trial (RCT) provides strong evidence that early surgery (within about 12 weeks) results in faster relief of pain and disability compared with conservative treatment.
• Economic analysis found that, when societal costs such as loss of earnings were taken into account, the difference in cost was negligible. However, the National Institute for Health and Clinical Excellence does not take loss of earnings into account in its economic analyses because this could bias resource allocation decisions against people who are not earning.
• The study also showed that, at the end of 1 year, outcomes were similar for the two approaches to management, and that dissatisfaction increased during the second year: 5% of people reported an unsatisfactory outcome at the end of the first year, and this increased to 20% the end of the second year.

Referral for physiotherapy

• Referral for physiotherapy is recommended for people who are recovering slowly because there is consistent evidence from three RCTs that physiotherapy may provide statistically significant and clinically important benefits for people with sciatica for at least a year. A Dutch economic analysis of one of the RCTs found that physiotherapy was not cost effective, but it is not clear if the conclusions are generalizable to the UK.

Referral for epidural injection of corticosteroids

• Considering referral for assessment for epidural injection of corticosteroid is recommended because this is performed by some specialists. However, a systematic review of RCTs and two additional RCTs provide mixed evidence, with some positive results and some negative results, on the benefits of epidural or extradural corticosteroid injection compared with placebo [Valat et al, 2003; Ng et al, 2005; Luijsterburg et al, 2007a]. Meta-analysis was not performed, because the studies were too heterogeneous.

These recommendations are developed from national guidelines and are largely based on expert opinion [NICE, 2001; Koes et al, 2007; Health Education Board for Scotland and Health and Safety Executive, 2008].