NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
National Guideline Centre (UK). Low Back Pain and Sciatica in Over 16s: Assessment and Management. London: National Institute for Health and Care Excellence (NICE); 2016 Nov. (NICE Guideline, No. 59.)
Guideline updates: December 2020: in the recommendation on stopping opioid analgesics NICE added links to other NICE guidelines and resources that support discussion with patients about opioid prescribing and safe withdrawal management. September 2020: NICE's original guidance on low back pain and sciatica in over 16s was published in 2016. It was partially updated in September 2020. See the NICE website for the guideline recommendations and the evidence review for the 2020 update. This document preserves evidence reviews and committee discussions for areas of the guideline that were not updated in 2020.
24.1. Introduction
The epidural space lies within the spinal canal, outside the dura mater, and contains the spinal nerve roots, fat, connective tissue and blood vessels. An epidural injection is an injection of a therapeutic substance into this canal. Administration may involve a caudal injection at the base of the spine, in the midline between the vertebral laminae (interlaminar epidural) or laterally, through the intervertebral foramen (transforaminal epidural, nerve root injection, dorsal root ganglion injection).
The most commonly used epidural injectate for the management of sciatica is corticosteroid, with or without local anaesthetic. The immunosuppressant and anti-inflammatory effects of corticosteroids provide a theoretical basis and rationale for epidural injection. However, some studies suggest that local anaesthetic epidural injection alone may also be therapeutic. Recent studies have also examined the role of anti-TNF (Tumour Necrosis Factor) agents into the epidural space on the premise of a TNF-α mediated inflammatory mechanism.
Although performed widely since the 1950s, the administration of steroids into the epidural space remains unlicensed. HES data from 2010–2011 estimates that nearly 79,000 epidural and nerve root injections were performed in England.125
Currently there are areas of uncertainty beyond the effectiveness of epidural injections to be considered, including the ideal route of administration, the use of imaging to improve accuracy, the timing of injection and the safety profile.
24.2. Review question: What is the clinical and cost effectiveness of epidural injections in the management of people with sciatica?
For full details see review protocol in Appendix C.
Table 30PICO characteristics of review question
| Population | People aged 16 or above with sciatica and:
|
|---|---|
| Interventions |
|
| Comparisons |
|
| Outcomes | Critical
|
| Study design | RCTs and SRs will be included in the first instance. If insufficient RCT evidence to form a recommendation is found, non-randomised studies will be included. |
24.3. Clinical evidence
24.3.1. Clinical evidence summary: Image guided epidurals
Twenty RCTs were included in the image-guided epidurals part of the review, of which 3 were reported in 7 studies, giving a total of 24 studies; these are summarised in Table 31 below.2,11,25,26,42,43,52,53,56,69,70,82,84,100,103,105,109,112,114,121,126,142,143,164 Karppinen 2001 was also reported in Karppinen 2001A, Manchikanti 2008 was also reported in Manchikanti 2012B and 2012I and Riew 2000 was also reported in Riew 2006. Evidence from these studies is summarised in the GRADE clinical evidence profile/clinical evidence summary below. See also the study selection flow chart in Appendix E, study evidence tables in Appendix H, forest plots in Appendix K, GRADE tables in Appendix J and excluded studies list in Appendix L.
Table 31
Summary of studies included in the review: image- guided.
There was no RCT data that could be incorporated in this review for the comparison of steroid versus placebo/sham. The search was therefore widened to look for cohort study data for this comparison; however no relevant cohort studies were identified. A combined search for the epidurals injections for sciatica review and the spinal injections review identified four Cochrane reviews 31,162,163,174. One of them163 was not included as it included studies in people with neuropathic pain syndromes and not low back pain. The others reviews31,162,174 were not included as the stratification of people with low back pain, low back pain with or without sciatica and sciatica was unclear, however any relevant studies included in the reviews were included and re-extracted in this review where appropriate. The studies included in these Cochrane reviews were individually assessed and included if they matched the protocol.
24.3.1.1. Heterogeneity
For the comparison of steroid and anaesthetic versus anaesthetic (>70% prolapse), there was substantial heterogeneity between the studies when they were meta-analysed for pain, responder criteria for pain (>50% reduction in pain) at both short and long term follow-ups, and for responder criteria for function at less than 4 months. Pre-specified subgroup analyses, by route of administration were performed on these outcomes which mostly explained the heterogeneity for pain at longer term follow-up and responder criteria for pain at both time points. Heterogeneity remained for pain and responder criteria for function at less than 4 months however. A random effects meta-analysis was therefore applied to these 2 outcomes, and the evidence was downgraded for inconsistency in GRADE.
For the comparison of steroid and anaesthetic versus anaesthetic (mixed population / unclear spinal pathologies), there was substantial heterogeneity between the studies when they were meta-analysed for pain at and function (ODI) at less than 4 months. Pre-specified subgroup analyses were performed on these outcomes however it did not explain the heterogeneity for pain, and could not be applied to because all of the studies used the same route of administration. A random effects meta-analysis was therefore applied to these 2 outcomes, and the evidence was downgraded for inconsistency in GRADE.
24.3.2. Clinical evidence summary: Non image guided epidurals
Fifteen RCTs were included in the review; these are summarised in Table 32 below.1,16,21,29,30,37,81,85,141,146,161,170 Evidence from these studies is summarised in the GRADE clinical evidence profile/clinical evidence summary below. All the studies compared non image guided epidurals of either steroid, anaesthetic agents, or a combination of both.
Table 32
Summary of studies included in the review: non image- guided.
No studies comparing the use of anti TNF were identified.
See also the study selection flow chart in Appendix E, study evidence tables in Appendix H, forest plots in Appendix K, GRADE tables in Appendix J and excluded studies list in Appendix L.
24.3.3. Data unsuitable for meta-analysis
Table 33. Image-guided steroid + anaesthetic versus usual care lumbar spinal stenosis
24.3.4. Clinical evidence summary: Image-guided epidurals
24.3.5. Clinical evidence summary: non-image-guided epidurals
24.4. Economic evidence
Published literature
Two economic evaluations were identified that included epidural injections for sciatica as a comparator and have been included in this review.141,159 These are summarised in the economic evidence profiles below (Table 58, Table 59) and the economic evidence table in Appendix I.
Table 58
Economic evidence profile: Steroid plus local anaesthetic (non-image guided) versus placebo.
Table 59
Economic evidence profile: Steroid (non-image guided) epidural versus usual care.
Five economic evaluations were selectively excluded due to a combination of applicability and methodological limitations.40,93,137,160,169 These studies are listed in Appendix M, with reasons for exclusion given.
See also the economic article selection flow chart in Appendix F.
The study by Spijker-Huiges 2015160 was not combined with the previous one as the costs were reported only from a societal perspective and the QALYs calculated did not match with the results of the previous study and the individual SF36 scores reported for each intervention, ie while the individual SF36 scores show an improvement in the group receiving epidural, the QALY estimates were in favour of the control group.
24.5. Evidence statements
24.5.1. Clinical
24.5.1.1. Image guided epidurals versus sham/placebo (primarily caused by >70% disc prolapse)
In people with sciatica there was clinical benefit of an anti-TNF epidural compared with placebo for leg pain demonstrated in evidence from 1 study at up to 4 months (very low quality, n=37). When the epidural was a steroid combined with an anaesthetic, there was clinical benefit favouring the intervention arm for leg pain and number of responders with greater than or equal to 50% reduction in pain, but no difference for function (n=65, moderate quality).When anaesthetic only was administered, no difference between anaesthetic or sham was observed for pain or number of responders. No evidence was available for the other critical outcomes or for steroid mono-therapy.
24.5.1.2. Image guided epidurals versus active control
In people with sciatica primarily caused by >70% prolapse, non-disc lesion, or unclear spinal pathologies, there was no clinical benefit of a steroid plus anaesthetic epidural compared with anaesthetic alone for pain and function at either short or longer term follow up (only data for up to 4 months was available for the unclear spinal pathologies evidence). The evidence ranged from very low to high quality, and from 1 to 4 studies, n=69 to 606. When anti-TNF was combined with anaesthetic (in sciatica primarily caused by >70% prolapse), there was no benefit compared to anaesthetic alone observed for pain or function at ≤4 months (1 study, low and moderate quality, n=56). No evidence was available for the other critical outcomes or for other interventions.
In people with sciatica primarily caused by >70% prolapse, there was clinical benefit of a steroid combined with anaesthetic epidural compared with combinations of non-invasive interventions or compared with anti-TNF with anaesthetic for leg pain or function at less than or equal to 4 months (1 study, moderate quality, n=100 and n=54 for the different comparisons respectively). There was also clinical benefit for quality of life for the comparison with non-invasive combinations. No evidence was available for the other critical outcomes or interventions.
24.5.1.3. Non image guided epidurals versus sham/placebo
In people with sciatica primarily caused by >70% prolapse, there was no clinical benefit of a steroid compared with placebo for function at greater than 4 months follow-up (low quality, 2 studies, n=221), and pain at up to 4 months (moderate quality, 2 studies, n=174). There was no evidence for this comparison for any of the critical outcomes in the population with an unclear pathology. When steroid was combined with anaesthetic (in sciatica with an unclear pathology) there was no clinical benefit for pain or function demonstrated by 1 study at both short and long term follow-ups (moderate and low quality, n=228).
24.5.1.4. Non image guided epidurals versus active control
In people with sciatica with an unclear pathology, there was a clinical benefit of steroid compared to usual care for leg pain and function demonstrated in evidence from 1 study at up to 4 months but not at greater than 4 months and most of the quality of life domains at both short and long term follow up (low quality, n=63).
In people with sciatica primarily caused by >70% prolapse, there was no clinical benefit of steroid combined with anaesthetic compared with pharmacological treatment (NSAIDs) for pain and function demonstrated in evidence from 1 study at up to 4 months (low quality, n=64), or for pain when compared with a combination of pharmacological interventions at both short and long term follow up (1 study, low and very low quality, n=50).
In people with sciatica primarily caused by >70% prolapse, there was clinical benefit of steroid combined with anaesthetic compared with anaesthetic demonstrated in evidence from 1 study for pain at up to 4 months when using a combination of methylprednisolone or triamcinolone in combination with bupivacaine (moderate quality, n=105). However there was no benefit when dexamethasone and bupivacaine were used (moderate quality, n=105). There was no evidence for any of the critical outcomes for this comparison in the sciatica caused by spinal stenosis or unclear pathology populations.
24.5.2. Economic
- One cost-utility analysis found that non-image guided epidural injections of steroid plus anaesthetic was not cost effective compared to placebo for adults with low back pain and sciatica (ICER: £44,701 per QALY gained). This analysis was assessed as partially applicable with potentially serious limitations.
- One cost-effectiveness analysis found that non-image guided steroid epidural was more costly and more effective than placebo for adults with sciatica (ICER: £60 per 1 point improvement in NRS back pain score). This analysis was assessed as partially applicable with potentially serious limitations.
24.6. Recommendations and link to evidence
| Recommendations |
|
|---|---|
| Research recommendations |
|
| Relative values of different outcomes | The GDG agreed that health related quality of life, pain severity, function and psychological distress were the outcomes that were critical for decision making. Adverse events (mortality and morbidity), and healthcare utilisation were also considered as important. For image-guided epidurals, evidence was reported for all of the critical outcomes, but there were limited data for quality of life. For non-image-guided epidurals there was no evidence for quality of life. |
| Trade-off between clinical benefits and harms | The GDG agreed that there was sufficient RCT evidence for all comparisons except for image-guided epidurals versus placebo/sham. However, there was no relevant cohort data found to address this. The GDG agreed that the evidence for the effectiveness of epidurals was conflicting. They noted that sciatic symptoms usually improve over the course of a few months in the majority of people without treatment. The placebo-controlled trials did show some evidence of an effect for epidurals, particularly for the combination of steroid plus anaesthetic. The evidence suggested the important component was the steroid, but there was no evidence of benefit of steroid alone or anaesthetic alone when compared to placebo/sham for any critical outcome. As a responder analysis suggested a 35% increase in the probability that people obtain substantial pain relief following epidural injections (steroid plus anaesthetic) compared to placebo, it was agreed that epidural injection of local anaesthetic and steroid should be considered as a treatment option. Most of the RCT evidence in the review came from people with acute and moderately severe sciatica, and the GDG considered that this would be the population most likely to benefit from epidural injection. The group discussed the evidence for anti-TNF. There was no evidence found for non-image guided anti-TNF, but there was evidence for image-guided anti-TNF epidurals. Despite the evidence showing a positive effect of image-guided anti-TNF epidurals on pain and function, the GDG noted that the evidence was limited as it came from three studies which could not be pooled together because different comparisons were used. The group discussed the risks associated with the different routes of administration of an epidural. The opinion of the group was that serious complications are very rare. The most common adverse event was a temporary increase in pain which the GDG considered could be outweighed by the potential benefits. The group discussed that there is some guidance in the UK suggesting epidurals should be given under image-guidance based on safety grounds, although there was limited evidence for a difference in effectiveness of image guided compared to non-image guided epidural injections from this review. It was therefore agreed that a recommendation for future research should be drafted to ascertain the evidence base for safety and effectiveness for image guided and non- image guided epidural injections. Summary Overall, the GDG considered that epidural injection, whether administered under image guidance or without, is a relatively safe and routinely used procedure, and had some evidence demonstrated by placebo-controlled trials for effectiveness in pain relief for epidurals of local anaesthetic and steroid. There was insufficient/ lack of evidence for effectiveness to support epidural injections using anti-TNF. The studies were conducted in small populations who had at least moderately severe sciatica and did not have further treatment options available to them (other than surgery). The evidence reviewed by the GDG suggests that epidural injection of local anaesthetic and steroid may reduce the number of people who would require surgical intervention. This evidence was reinforced by evidence from 2 trials that were included in the spinal decompression review (See Chapter 28) that compared decompression to epidurals showing that 50% of people who had an epidural did not go on to have surgery. The group therefore agreed that in acute, severe sciatica where patients would otherwise be offered surgery, an epidural injection of local anaesthetic and steroid should be considered. The group discussed the evidence that had been conducted in sciatica patients with central spinal canal stenosis. The populations studied comprised people with neurogenic claudication primarily. There was insufficient evidence that epidural injections of local anaesthetic and steroid were effective in this group of people and it was noted that current opinion also reflects this. The group therefore agreed to make a recommendation against using epidurals in people with claudicant leg symptoms caused by central spinal canal stenosis. The GDG discussed that the purpose of this review had been to determine efficacy of different injectates, rather than comparing image guided to non-image guided injections. However, the stratification of the review by those delivered under image guided to those that weren't did not demonstrate a clear indication of improved efficacy of image guided epidurals over non-image guided. They therefore agreed that a research recommendation was warranted in this area. |
| Trade-off between net clinical effects and costs | Two economic evaluations were included comparing non-image guided epidural with either usual care or placebo in a population of adults with sciatica.141, 159 In particular the study by Price et al. (2005)141 was a cost-utility analysis comparing the intervention with placebo which concluded that epidural increased costs and improved health (increased QALYs), with an incremental cost-effectiveness ratio of £44,701 per QALY gained. In sensitivity analysis where the costs were adjusted assuming only 1 epidural injection was administered and the impact on QALYs was assumed to be unchanged, the ICER went down to £25,746. The group noted that as the recommendation for epidurals was for the acute sciatica population (most likely to be defined as having symptoms for <3 months), then multiple injections would not usually be performed within this short period of time. The GDG discussed the likely higher effectiveness observed with placebo as opposed to no treatment and concluded that if epidural was compared to no treatment or usual care it would probably be associated with a higher QALY gain, and therefore it would be more cost effective. In the same study no cost was attached to the placebo arm while in reality patients could incur the cost of other treatments such as medications and the cost of their side effects. The GDG noted that the studies from which the cost effectiveness data was derived did not have a diagnosis of sciatica confirmed by imaging. The GDG felt that clinical diagnosis alone may overestimate the numbers of patients with true sciatica and lower their confidence in the results. There was evidence suggesting that epidural injection may reduce the number of people with severe sciatica requiring surgical intervention; this would generate some cost savings. For these reasons, the GDG decided not to make a strong recommendation on epidural injections but they concluded that they may be cost effective for some patients and therefore it should be considered. |
| Quality of evidence | The quality of the evidence for both image guided and non-image guided epidurals was mostly low or moderate (due to risk of bias usually caused by selection or performance bias, small sample sizes and imprecision) across all of the outcomes and comparisons in the review. There was evidence to show an effect of anti-TNF (image-guided), however this was only from single studies, which mostly had small sample sizes. Some of the studies had incomplete reporting of outcome data (for example, no standard deviations were reported for some outcomes and 1 study only had data for 1 participant in the comparison arm). This also meant that the evidence was rated as being at high risk of bias and so overall the group did not have confidence in the findings. The GDG had more confidence in the evidence for epidurals in sciatica patients with spinal stenosis (steroid was given as an adjunct) because the main study contributing to the meta-analysis was conducted in 400 participants. The group were less confident in the results of the other contributing study, since it was smaller and although it was also conducted in spinal stenosis patients, it differed considerably to the other studies in the review. The population consisted of chronic sciatica patients with over 100 months of pain, and patients could be given as many epidural injections as they needed (the average given was 4). The GDG felt that this did not reflect clinical practice. |
| Other considerations | The group discussed the effectiveness of giving multiple / subsequent epidural injections. The group noted that as the recommendation for epidurals was for the acute sciatica population (most likely to be defined as having symptoms for <3 months), then multiple injections would not usually be performed within this short period of time. The GDG agreed that this recommendation would equally apply for pregnant women and should be considered alongside BNF guidance. The GDG were aware of existing NICE interventional procedure guidance for Therapeutic endoscopic division of epidural adhesions (IPG333) recommending special arrangements for clinical governance, consent, audit and research.124 This procedure was therefore excluded from this review and if it's use is considered for people with sciatica, existing guidance should be followed. Research recommendation Why this is important: Epidural injection of therapeutic substances that include corticosteroids is commonly offered to people with sciatica. Epidural injection might improve symptoms, reduce disability and speed up return to normal activities. Several different procedures have been developed for epidural delivery of corticosteroids. Some practitioners inject substances through the caudal opening to the spinal canal in the sacrum (caudal epidural), whereas others direct the injection through the foraminal space at the presumed level of nerve root irritation (transforaminal epidural). There is a rationale that transforaminal epidurals might be most effective, by ensuring delivery of corticosteroids directly to the region in which the nerve root might be compromised. However, transforaminal epidural injection requires imaging, usually within a specialist setting, potentially limiting treatment access and increasing costs. Caudal epidural injection might be undertaken without imaging, or with ultrasound guidance in a non-specialist setting, but, it has been argued, the drug might not reach the affected nerve root and therefore this approach might not be as effective as would be transforaminal injection. Empirical evidence that 1 approach is clearly superior to the other is currently lacking. Access to the two procedures varies between healthcare providers, and people who do not respond to caudal corticosteroid injection might subsequently receive image guided epidural injection. People with sciatica might therefore currently experience unnecessary symptoms at unnecessary cost to the NHS than would be the case if the most cost effective modes of delivering epidural corticosteroid injections were used. |
- Epidural injections for sciatica - Low Back Pain and Sciatica in Over 16s: Asses...Epidural injections for sciatica - Low Back Pain and Sciatica in Over 16s: Assessment and Management
- earP [Pseudomonas putida KT2440]earP [Pseudomonas putida KT2440]Gene ID:1043242Gene
- Gm14856 predicted gene 14856 [Mus musculus]Gm14856 predicted gene 14856 [Mus musculus]Gene ID:671297Gene
Your browsing activity is empty.
Activity recording is turned off.
See more...