Epidural steroid injections for low back pain

BMJ 2011; 343 doi: http://dx.doi.org/10.1136/bmj.d5310 (Published 13 September 2011) Cite this as: BMJ 2011;343:d5310
  1. Steven P Cohen, associate professor
  1. 1Departments of Anesthesiology and Critical Care Medicine, Johns Hopkins School of Medicine, Baltimore, MD 21029, USA
  1. scohen40{at}jhmi.edu

Overall the evidence of benefit is weak, but some may benefit

In the linked randomised trial (doi:10.1136/bmj.d5278), Iversen and colleagues assess the efficacy of caudal epidural steroid and caudal epidural saline injection in the treatment of chronic (>12 weeks’ duration) lumbar radiculopathy.1 Low back pain is the leading cause of disability in the world and a major contributor to a wide range of other problems, such as substance misuse and depression.2 Perhaps the most important distinction to make when categorising low back pain is to distinguish between neuropathic and mechanical causes, because this informs treatment.2 3 Recent studies show that about a third of patients with chronic low back pain have predominantly neuropathic pain.3 Sciatica responds better than non-specific low back pain to interventions, but treating any form of low back pain is challenging.

Epidural steroid injections have been used for more than 50 years to treat low back pain and are the most common intervention in pain clinics throughout the world.4 Yet despite their widespread use, their efficacy is unclear. Of around 35 controlled studies evaluating such injections, slightly more than half show some benefit. Moreover, systematic reviews written by authors who perform epidural steroid injections are more likely to conclude that it is effective than reviews written by those who do not use such injections.5 6 7 Reasons for these discrepancies include differences in methodology, treatment characteristics, and perhaps most importantly, patient selection.

Iversen and colleagues’ study had three intervention groups, each of which received two injections with a two week interval: subcutaneous sham injections superficial to the sacral hiatus and not into the spinal canal; caudal epidural injections with saline alone; and caudal epidural injections with a combination of saline and triamcinolone acetonide. All groups improved after the intervention, but no significant difference was seen in the primary outcome (as measured by the Oswestry disability index) between the groups over time.

The decision to use a second control group that received epidural saline was a shrewd addition because evidence suggests that the epidural administration of non-steroid solutions may provide benefit.7 This may be due to the washout of inflammatory cytokines, lysis of scar tissue, and the effects of local anaesthetic. Yet this strategy, designed to distinguish between the effects of perineural steroids and the epidural route itself, is futile if a treatment benefit is not shown, as was the case for Iversen and colleagues’ study. A similar study found that epidural steroid injections provided better pain relief than either epidural saline or intramuscular steroids, which in turn were both more effective than intramuscular saline.8

How can these differences be explained? Overall, epidural steroid injections seem to be beneficial, but only provide modest improvement in carefully selected patients with predominantly radicular symptoms. There are several potential reasons why Iversen and colleagues’ study failed to show a benefit. Firstly, the authors injected steroids via the caudal root, which means that the area(s) they were targeting (for example, L4-S1) were far removed from the site of injection. Although caudal injections can be effective for neuropathic back pain, site targeted injections that deposit the steroid directly over the affected nerve root and into the ventral epidural space (transforaminal injections) are superior.5 Secondly, part of the effect of caudal epidural steroid injections stems from the large volume injected, which increases the likelihood of spread to the area of pathology, and may itself afford benefit.9 However, any benefits conferred by this were negated by the combination of the low dose of steroid (40 mg triamcinolone) used and high volume (30 mL), which significantly diluted the steroid bathing the nerve roots. In addition, most studies use a local anaesthetic to break the cycle of pain, increase blood flow to ischaemic nerve roots, and possibly reverse the processes of central sensitisation, but the authors did not do this. A final shortcoming lies in the randomisation, which resulted in the sham group having experienced a significantly shorter duration of pain. Studies conducted in multiple populations with back pain have found an inverse correlation between the duration of symptoms and treatment outcomes.2 10

So where do we go from here? After around 35 studies have failed to provide a definitive answer regarding the efficacy of epidural steroid injections, it is unlikely that future trials will do so. This poses a dilemma for researchers designing clinical studies for low back pain. Generally, efficacy studies command stringent inclusion and exclusion criteria (such as short duration of pain, less disease burden), which maximises the chances for pain relief. However, these conditions also increase the likelihood of the control group improving. The placebo response is stronger for pain than for almost any other non-psychiatric medical condition. Whereas the response of the control groups can be reduced by liberalising the selection criteria, the treatment response will invariably also be reduced. More inclusive selection criteria tend to be used for comparative-effectiveness studies because they enhance external validity, but this will in turn reduce the chance of demonstrating efficacy. For a study with relatively liberal selection criteria that used a less effective injection route, Iversen and colleagues’ study was probably underpowered.

Despite the negative findings, the current study should not be misinterpreted as suggesting that epidural steroid injections are of no use in neuropathic back pain. Even if only a small proportion of people return to work11 or can avoid surgery—as a randomised study performed by spine surgeons found12—this suggests that epidural steroid injections may be an effective adjunct when used judiciously.

Although these findings do not provide a definitive answer regarding the effectiveness of epidural steroids, they do fit neatly into an increasingly complex puzzle. In patients with acute or subacute radiculopathy secondary to a herniated disc in whom more conservative measures have failed, epidural steroid injections should be considered as part of a multidisciplinary treatment plan (including exercise and physiotherapy). People with chronic or non-remitting pain, non-radicular pain, and spinal stenosis may also benefit, but the number needed to treat is considerably higher. Fluoroscopy should always be used to ensure proper needle placement, and the transforaminal approach seems to provide better relief than the caudal or conventional interlaminar method.


Cite this as: BMJ 2011;343:d5310


  • Research, doi:10.1136/bmj.d5278
  • Competing interests: The author has completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declares: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned; not externally peer reviewed.