Overview

Low back pain without disc herniation is the most common problem among chronic pain disorders, but a patho-anatomical cause can be established in only 15 % of all cases.¹ Treatments to relieve this affliction have been many, among them spinal injections – engendering much controversy in the literature; opinions about efficacy, safety and relevance have differed greatly since their inception in the 1920s, with many studies considered poor quality.^2–17

Although epidural injections are one of the most commonly used invasive interventions in the treatment of low back pain, with or without radicular pain, there is currently little consensus about this technique and wide variation in practice.²¹ There is also no agreement on the most effective approach for lumbar epidural injection, whether to use steroid, local anaesthetic, saline or a combination, or the exact volume required. Depot steroids are not licensed for spinal use^18,19 but orthopaedic and pain specialists, rheumatologists and others use these injections extensively.²⁰ The caudal route of administration may require a larger volume but is least likely to cause dural puncture.^22,23

A paucity of well designed, randomized controlled studies, and a lack of statistically significant results in the existing literature mean that a solid foundation for the effectiveness of spinal injection therapy is lacking.⁹ NICE, the UK National Institute for Health and Clinical Excellence, recommended that patients with persistent non-specific low back pain should not be offered injections of therapeutic substances,²⁴ but what impact this has had on clinical practice is uncertain.

A Cochrane Review found minor side-effects such as headache, dizziness, transient local pain, tingling, numbness and nausea reported in a small number of patients in only half the trials reviewed. The review concluded that there is no strong evidence for or against the use of any type of injection therapy for individuals with subacute or chronic low-back pain.¹⁰

Safety

All the contraindications listed in Section 2 apply, but particularly:

The incidence of intravascular uptake during lumbar spinal injection procedures is approximately 8.5%; it is greater in patients over 50, and if the caudal route is used rises to 11%. Absence of flashback of blood on pre-injection aspiration does not predict extravascular needle placement.³¹ Epidural steroid injection is safe in patients receiving aspirin-like antiplatelet medications, with no excess risk of serious haemorrhagic complications, i.e. spinal haematoma. Increased age, large needle gauge, needle approach, insertion at multiple interspaces, number of needle passes, large volume of injectant and accidental dural puncture are all relative risk factors for minor haemorrhagic complications.³²

Safety precautions and strict aseptic techniques are the same as for all injections. An additional hazard is the rare possibility of an intrathecal injection of local anaesthetic which may be avoided by using corticosteroid alone. The rationale is that the benefit of the brief relief of pain and the diagnostic information obtained from using an anaesthetic does not outweigh the potential risks. Normal saline can be added or Adcortyl used instead of Kenalog if additional volume is required.

New neurological symptoms or worsening of pre-existing complaints that persist for more than 24 hours (median duration of symptoms 3 days, range 1–20 days) might occur after epidural injection,³² but in the authors’ experience this is rare.

The British Society for Rheumatology and the Royal College of Anaesthetists produce guidelines for the use of epidural injections. We commend them to all practitioners who give these injections. They can be found at:

Accuracy

Performing spinal injections under imaging can ensure correct placement but requires specialized training and is expensive to perform, especially if done in theatre; many doctors perform these techniques ‘blind’ and obtain satisfactory results.

Accuracy of blind caudal epidural injections compared with targeted placement has been assessed in a few studies. In one, successful placement on the first attempt occurred in three out of four subjects. Results were improved when anatomical landmarks were identified easily (88%) and no air was palpable subcutaneously over the sacrum when injected through the needle (83%). The combination of these two signs predicted a successful injection in 91% of attempts. In another study blind injections were correctly placed in only two out of three attempts, even when the operator was confident of accurate placement. When the operator was less certain, the success rate was less than half and if the patient was obese the success rate reduced even further. In a third prospective randomized, double-blind trial, the results showed no advantage of spinal endoscopic placement compared with the more traditional caudal approach.26–29,^34,39

Efficacy

Lumber epidurals: a systematic review of epidural corticosteroids for back pain found at least 75% pain relief in the short term (1–60 days) with the number needed to treat (NNT) of 7 (7–16) and at least 50% pain relief in the long term (3–12 months) with NNT of 13 (7–314).³ A randomized, double-blind, controlled trial concluded that lumbar interlaminar epidural of local anaesthetic with steroid was effective in 86% of patients, and without steroid in 74%.³¹

A systematic review indicated positive evidence (Level II-2) for short-term relief of pain from disc herniation or radiculitis utilizing blind interlaminar epidural steroid injections; there was less strong evidence for long-term pain relief for these conditions and for the short- and long-term relief of pain from spinal stenosis and from discogenic pain without radiculitis or disc herniation.²⁵ Another review of both caudal and lumbar epidurals also concluded that the best studies showed inconsistent results and benefits were of short duration only.⁶ Yet another showed strong evidence for epidurals in the management of nerve root pain due to disc prolapse, but limited evidence in spinal stenosis.²² A multicentre randomized controlled trial of epidurals for sciatica reported significant relief at 3 weeks but no long-term benefit.¹⁴

In the past, large volumes have been injected into the epidural space;³³ however, a total injection volume of 8 ml is sufficient for a caudal epidural injection to reach the L4/5 level.³⁴

Selective guided nerve-root injections of corticosteroids are significantly more effective than those of bupivacaine alone in obviating the need for operative decompression for 13–28 months following the injections in operative candidates. This finding suggests that patients who have lumbar radicular pain at one or two levels should be considered for treatment with selective nerve-root injections of corticosteroids prior to operative intervention. A significantly greater proportion of patients treated with transforaminal injection of steroid achieve relief of pain compared with those treated by transforaminal injection of local anesthetic or saline or intramuscular steroids.³⁰ When symptoms have been present for more than 12 months, local anaesthetic alone may be just as effective as steroid and local anaesthetic together.

When conservative measures fail, nerve-root injections are effective in reducing radicular pain in patients with osteoporotic vertebral fractures and no evidence of nerve root palsy. These patients may be considered for this treatment before percutaneous vertebroplasty or operative intervention is attempted.^35,36,38

Injection of the sacroiliac joints for painful sacroiliitis appears to be safe and effective. It can be considered in patients with contraindications or complications with NSAIDs, or if other medical treatment is ineffective,³⁷ though often manipulative techniques can obviate the need for an injection. However, accurate placement of the drug without the use of fluoroscopy is estimated to be successful in only 12 % of patients.⁴⁰

Indications for spinal injection

The techniques described here include caudal epidural, nerve root, facet joint, sacroiliac joint and sacrococcygeal joint injections and the far less common technique for cervical nerve root pain. The choice between giving a caudal or nerve root injection can be aided by the site of pain; if this is clearly unilateral in the lumbar area, or radiating down one leg, a nerve root injection may be effective. If the pain is bilateral or central in the lumbar spine, a caudal epidural may be a better choice; however, this guide is not an absolute.

The following are the main indications for caudal and nerve root injections:

Older patients with chronic back pain and stiffness increased on active extension may benefit from facet joint injections. A retrospective study of patients with spinal stenosis found that 35 % of patients had at least 50 % improvement; those with spondylolisthesis, single level stenosis and older than 73 had better outcomes.²⁸ Less commonly, injections for coccydinia or sacroiliac joint pain can be attempted in cases of acute traumatic or post-natal pain.

Summary

There is a wide variation of opinion about the efficacy of spinal injections for back pain; adverse effects are generally minor and it cannot be ruled out that specific subgroups of patients may respond to a specific type of injection therapy. A cost-effective intervention which may be performed safely as an outpatient procedure and rapidly relieve pain, even in the short term, is worth considering for carefully selected patients with both acute and chronic low back pain provided that, as with all injection techniques, resuscitation facilities are available and the guidelines on aseptic technique are strictly followed.