Who we are
Site map
Coming meetings
Meeting reports
Notice board
Pain topics
Slide shows
Case reports


Epidural Steroids And Quality Of Life

Dr  J C D Wells, Liverpool, UK

For a treatment to be useful in chronic pain, it must not only reduce the pain, but must show other characteristics. It must be relatively lacking in side-effects. It should be shown to improve other aspects of quality of life (QOL). It should be reasonably cost-effective.

QOL is not often measured in research on nerve blocks in general and in epidurals in particular. This is perhaps a significant deficiency, in that huge numbers of epidural steroid injections are administered every year. Further research needs to be done on the use of epidural steroids; this would not be particularly difficult to do and would be fertile ground for young doctors to carry out good quality research.

Concerning pain relief and QOL. Anaesthesia is not a useful treatment for chronic pain, because there is no quality of life for the sufferer. Death stops pain completely, hence both suicide and euthanasia have been used by patients and others for their chronic pain. However, obviously the resulting QOL is nil! Morphine, an excellent painkiller, has been under-used for both malignant and non-malignant pain because of fears of side-effects, which would reduce QOL.

Sicard and Cathelin described caudal epidurals in 1901 (Burkle, 2002). This work was repeated by Viner (1925) and in the early '30s in Chicago, lumbar segmental anaesthesia was described (Dogliotti, 1939). Evans (1930) described the use of procaine and saline; Robecchi first used hydrocortisone via the first sacral root canal (1952). Boudin (1955) and Lievre (1957) both published in Paris in the late '50s; they injected steroids intrathecally. The first good publication in the English language was by Goebert (1961). He described the use of Procaine and Hydrocortisone, and this gradually became the norm, depot steroids being introduced as they became available. Swerdlow (Manchester) also wrote an excellent paper (1970), studying 325 patients with sciatica. Numerous papers have been published since, not all useful.

What QOL measures can be considered, and what has been published in trials? We should consider both efficacy and safety.

Parameters to look at regarding efficacy include the following:-

  • Reduction in pain (how shall we measure this?)

  • Reduction in medication needs Improvement in mood

  • Improvement in function

  • Improvement in sleep pattern

  • Return to work

  • Reduced health care utilisation (including subsequent surgery in patients with disc prolapse)

  • Reduced benefit utilisation (ie, disability, ? sickness benefit, early retirement, etc)

Under "safety" we can consider side-effects, both minor and major, and complications.


History of Epidurals

James Corning, a New York City Neurologist, gave a thoracic epidural (T11/12) with cocaine (1885). The patient's legs became numb, with spread to the lumbar region, the penis and scrotum. However, the patient could still walk. The effects wore off within an hour.

When we look at the literature, this reveals significant discrepancies which render comparison between patient groups extremely difficult. These discrepancies include:-

  1. Sample description (eg, acute versus chronic)
  2. Cause of symptoms (prolapsed disc, spinal stenosis, epidural fibrosis or don't know!)
  3. Sample numbers (often very small, understandably)
  4. Variations in technique

Variations in technique are numerous. Several approaches are used - the lumbar, caudal and cervical - and there appears to be no difference between whether the lumbar or caudal route is used for low back pain. New techniques, such as Racz catheter, transforaminal or even epiduroscopy, are now producing well-researched results.

Different volumes of injectate are used for different routes and even the same route. The contents of the injectate vary, including local anaesthetic or saline mixed with steroid and other additives, including clonidine, adrenaline and ketamine. The steroid itself varies and there is no uniformly agreed dosage, nor indeed is any allowance ever made for the weight of the patient. Different types of needle are used, many Anaesthetists liking the Tuohy needle, but there being no rationale to use this for a single-shot injection. Finally, X-ray control, in my opinion mandatory, is often not used and is sometimes not reported.

Identification of level of block and correct needle placement are both notoriously unreliable, using a blind approach. Correct needle placement has been described in as few as 30 per cent and up to 75 per cent of patients only (Steward, 1987; White, 1980). Epidural fibrosis or adhesions (often occurring after surgery) may make spread to the affected nerve root difficult or impossible (Greenwood, 1952; Racz, 1997). Deposition of epidural steroids at the side of the pathology occurred in only 25 per cent of patients (Fredman, 1999).

Huge numbers of epidural steroid injections are performed every year. The worldwide numbers probably exceed a million. That means that since 1955 some 50 million have been carried out. The Quebec Task Force note that an intervention practiced commonly, but without significant scientific evidence, still constitutes the third strongest type of evidence for treatment efficacy and this is apparently where we are with epidural steroids (Hopwood, 1999; Agency for Health Care Policy and Research, 1994). However, the Agency for Healthcare Policy and Research in the States regards the quality of evidence as "C". That means that there is no good evidence, particularly from clinicial trials, to endorse efficacy. Are they correct?

The results of treatment have been described in over 6 thousand patients in 70 reports in the world literature. The studies can be considered in 4 groups:-

  1. Older studies
  2. Randomised controlled trials (RCTs)
  3. Systematic reviews and meta-analysis
  4. Recent studies

The older studies are important, because they popularised epidural steroids (ES) and made them accepted in clinical practice. However, they are usually small, often lack a control group and have major design flaws. They usually only consider pain relief and no other QOL measures. These include the studies by Goebert and Swerdlow.

The second group of RCTs yield diverse results. There is no standardisation of diagnosis and techniques between trials, so some of them are like comparing apples with oranges for taste. Several authors have performed systematic reviews or meta-analyses on these figures (Koes, 1999; Watts, 1995; McQuay, 1998).

Some of these papers look at QOL issues. For instance, Dilke in 1973 reported on 100 patients, all of whom had disc disease and unilateral sciatica with neurological deficit, and assessed not only reduction in pain, but also decreased usage of analgesics, return to work and surgery. All of these showed favourable benefit after ES, but there was no effect on neurological deficit or straight leg raising. Number needed to treat (NNT) was 3.5 at 60 days and 6.3 at 3 months.

A recent paper by Bowman et al (1993) looked at 35 patients. The methodology was poor; however, it did look at mood and disability, and both disability and pain were reduced at one week, with a maintenance of reduction of disability at 3 months. There was little statistical evidence of improvement of mood, but there was not a great deal of mood disturbance beforehand.

Buchner et al (2000) looked at 36 patients with radicular pain and measured VAS, SLR and functional status. All of these improved in the active treatment group after 2 weeks, but this effect had worn off by 6 weeks.

Valat et al (2003) reported on 85 patients and found not only a reduction in VAS, but an improved Roland Morris score and a significant reduction in anxiety and depression, although none of these measures reached statistical significance.

Vad et al (2002) considered VAS, Roland Morris and stretch testing, and treated 2 groups of patients, one with transforaminal steroids and one with trigger point injections of steroids. The transforaminal group showed an improvement in 84 per cent at 16 months; the control group 48 per cent.

An important paper by Aldret (2003) used 2 doses of Indomethacin epidurally compared with Methylprednisolone, and was a prospective comparative trial on patients who had radiculopathy after surgery. Two hundred and six patients were included. There was a reduction in VAS, increased physical activity, reduction in emotional distress and a reduced intake of medication in all 3 groups. Two milligrams of Indomethacin was better than 1 milligram, and equivalent to 80 mg of Methylprednisolone.

If we look at the RCTs included in the work of Koes, Watts and McQuay (Table I) it can be seen that there are huge differences between results. It is hard to believe that these are on identical patient populations, or carried out in the same manner. It is interesting that the French were the first authors to popularise the method, but now French papers seem to show no benefit. Other papers which show less efficacy were published by Orthopaedic Surgeons and Rheumatologists (? accuracy of injection).

Obviously this is not allowed, but if one takes the results of papers carried out by the Pain Specialist with anaesthetic training, it is interesting to note NNTs in the region of 3 for short-term relief and 6 for long-term relief. However, the actual figures looking at all papers are 7.3 for greater than 75 per cent pain relief up to 60 days, and 13 for greater than 50 per cent pain relief up to 1 year. McQuay also points out that there is a very high placebo or spontaneous response in many of the trials.

Studies are limited in cervical epidurals, and look only at pain reduction with no QOL measures. I have found less than 200 patients described in 4 studies, but all claim good results.


Complications and Side-effects

Abram and O'Connor (1966) looked at 53 series of descriptions of epidurals, numbering 66 thousand. They came up with only 40 complications, 37 taps, 1 congestive cardiac failure, 1 deep vein thrombosis and 1 skin infection. However, Watts and Silegy felt the likelihood of dural tap was about 2 and a half per cent (ie, 1 in 40).

Other complications in the literature include adrenal suppression, Cushing's syndrome, meningitis or meningism, abscess, long-term dural leak, air embolus and allergy.

There is a significant incidence of neurological sequelae after epidural injections. This can be due to direct trauma, the injectate (? steroids intrathecally) or blood supply to the spinal cord.

Side-effects. Mild side-effects include nausea, headache, dizziness, vasovagal attacks and flushing of the face. These are uncommon. Systemic effects of steroids include Cushing's syndrome, Cushingoid features (moon fascies, fat deposition, fluid retention and skin lesions), adrenal suppression (2 to 4 weeks), elevated cortesol level (2 weeks), increased blood glucose concentration (special attention in diabetics) and menstrual irregularities.

Less common side-effects include unintentional dural puncture, with the potential for post-spinal tap headache (dependent on age and other factors) and also the potential for sub-arachnoid injection, which could be disastrous. Even less common, however, are epidural haematomas, infection including abscess or discitis, arachnoiditis (additives) or meningitis.

Long-acting preparations contain a preservative such as polyethyleneglycol and benzylalcohol, and it is inadvisable to administer these intrathecally. The only way that I know of preventing this is to use X-ray control and screening.


Should We Use Epidural Steroids?

It seems to me that there is good evidence for their use in patients with nerve root involvement, if we wish to produce short-term pain relief and combine this with rehabilitation. There needs to be an absence of contra-indications such as infection, coagulopathy or unstable diabetes. There needs to have been no response to conservative treatment but this will vary, according to the pain and disability of the patient. However, certainly at 6 weeks if the condition is not improving, there is less likelihood of it improving spontaneously. In patients with long histories of 6 months or more of chronic, radicular pain, treatment is obviously going to be less effective, but by using fluoroscopy or a Racz catheter or a transforaminal approach, good results can still be obtained in selected patients. There is little compelling evidence for the use of epidural steroids in chronic back pain only, although they can be combined with facet joint denervation in appropriate patients.

So we would look for patients with radicular pain, possibly from disc herniation, or patients with failed surgery, patients who are motivated with postural back pain with some radicular features and patients with cancer pain and tumour infiltration.

In the future, it is vital to consider further research on this subject. There is surprisingly little good scientific evidence from RCTs and very little on QOL. We know that there is a short-term effect in a significant number of patients and a long-term effect in a few. This benefit should be combined with rehabilitation to maximise outcome. We need to discuss with the patient the lack of a product licence for epidural steroids and also discuss with them the side-effects. My own feeling is that it is essential to use X-ray control, whenever this is obtainable, for best results.

In the future, further research needs to be done on QOL factors following epidurals including reduction of medication, improvement of mood, function and sleep pattern and social indicators such as return to work, reduction in health care utilisation and reduction in benefit utilisation. Only then can we convince our politicians and paymasters that this technique, which we use every day, is indeed not only effective but cost-effective.



Abram S, O'Connor T (1996): Reg Anaesth, 21(2): 149-162

Agency for Health Care Policy and Research (December 1994): Acute low Back Pain Problems in Adults. Clinical Practice Guidelines No 14. AHCPR Publication No 95-0642

Aldrete (2003): Anesth Analg; 96(2): 463-468

Beliveau P (1971): A Comparison between epidural Anesthesia with and without Corticosteroids in the Treatment of Sciatica. Rheumatol Phys Med 11: 40-43

Boudin G, Barbizet J, Guihard J (1955): L'Hydrocortisone intraarachnidienne, ses Applications cliniques en particulier dans le Traitement de la Meningite tuberculeuse. Bull Soc Med Hôp (Paris) 21: 817-821

Bowman et al (1993): In Spine Aug 1993; 18(10): 1345-1350

Breivik H et al (1976): Treatment of chronic low Back Pain and Sciatica. Comparison of caudal epidural Injections of Bupivacaine and Methylprednisolone with Bupivacaine followed by Saline. In Bonica J (Ed): Advances in Pain Research and Therapy, Vol 1, New York, Raven Press, pp 927-932

Buchner et al (2000): Clinical Orthopaedo Acclin, Orthop (375): 1492156

Burkle CM et al (2002): Beyond Blocks: The History of the Development of Techniques in Regional Anesthesia. In: Textbook of Regional Anesthesia (PP Raj, Ed), Ch 2, p 27

Bush K, Hillier S (1991): A controlled Study of caudal epidural Injections of Triamcinolone plus Procaine for the Management of intractable Sciatica. Spine 16: 572-575

Carette S et al (1997): Epidural Corticosteroid Injections for Sciatica due to herniated Nucleus Pulposus. N Engl J Med 336: 1634-1640

Corning JL (1885): Spinal Anaesthesia and local Medication of the Cord. N Y Med J 42: 183-185

Cuckler JM et al (1970): The Treatment (? use/efficacy) of epidural Steroids in the Treatment of radicular Pain. Anesthesia 25: 346-348

Dilke TFW et al (1973): Extradural Corticosteroid Injection in the Management of lumbar Nerve Root Compression. BMJ 2: 635-637

Dogliotti AM (1939): Anesthesia, Chicago, SB Debour, pp 529-544

Evans W (1930): Intrasacral epidural Injection in the Treatment of Sciatica. Lancet 2: 1225-1229

Fredman B, Nun MB, Zohar E (1999): Epidural Steroids for treating "failed Back Surgery Syndrome"; Is Fluoroscopy really necessary? Anesth Analg 88:367-372

Goebert HW, Jallo SJ, Gardner WJ et al (1961): Painful Radiculopathy treated with epidural Injections of Procaine and Hydrocortisone Acetate: Results in 113 Patients. Anesth Analg 40: 130-134

Greenwood J, McGuire TH, Kimbell F (1952): Study of the Causes of Failure in herniated intervertebral Disc Operation; Analysis of 67 re- operated Cases. J Neurosurg 9: 15-20

Hopwood MB (1999): Lumbar Epidural Steroid Injections: Is a clinical Trial necessary or appropriate? Reg Anesth Pain Med 24: 5-7

Klenerman L et al (1984): Lumbar epidural Injections in the Treatment of Sciatica. Br J Rheumatol 23: 35-38

Koes BW, Scholten RPM, Mens JMA, Bouter LM (1995): Efficacy of epidural Steroid Injections for low Back Pain and Sciatica: A systematic Review of randomized clinical Trials. Pain 63: 279-288

Koes BW, Scholten JPM, Mens JMA et al (1999): Epidural Steroid Injections for low Back Pain and Sciatica: An updated systematic Review of randomized clinical Trials. Pain Digest 9: 241-247

Lievre JA, Block-Michel H, Attali P (1957): L'Injection transsacrée Etude clinique et radiologique. Bull Soc Med Hôp (Paris) 73: 1110-1117

McQuay HJ, Moore RA (1998): An Evidence-based Resource for Pain Relief. London, Oxford University Press

Mathews JA et al (1987): Back Pain and Sciatica: Controlled Trials of Manipulation, Traction, sclerosant and epidural Injections. Br J Rheumatol 26: 416-423

Popiolek A et al (1991): Epidural Injections of Steroids in the Treatment of Patients with Sciatica in Discopathy. Neurol Neurochir Pol 25(5): 640-646

Racz GB, Heavner JE, Raj PP (1997): Epidural Neuroplasty. Semin Anesth 16: 302-312

Ridley MG et al (1988): Outpatient epidural Corticosteroid Injection in the Management of Sciatica. Br J Rheumatol 27: 295-299

Robecchi A, Capra R (1952): Lidocortisone (composto F): Prime Esperienze Cliniche in Campo reumatologico. Minerva Med 98: 1259-1263

Snoek W et al (1977): Double blind Evaluation of epidural Methylprednisolone for herniated lumbar Disc. Acta Orthop Scand 48: 635-641

Steward HD, Quinell RC, Dann N (1987): Epidurography in the Management of Sciatica. Br J Rheumatol 26: 424-429

Swerdlow M, Sayle-Creer W (1970): A Study of extradural Medication in the Relief of the lumbosciatic Syndrome. Anaesthesia 25: 341-345

Vad et al (2002): Spine, Vol 27; 1: 11-16

Valat et al (2003): Ann Rheum Dis 62: 639-643

Viner N (1925): Intractable Sciatica - the sacral epidural Injection - an effective Method of giving Relief. Can Med Assoc J 15: 630-634

Watts RW, Silagy CA (1995): A meta-Analysis on the Efficacy of epidural Corticosteroids in the Treatment of Sciatica. Anaesthesia and Intensive Care 23: 564-569

White AH, Derby R, Wynne G (1980): Epidural Injections for the Diagnosis and Treatment of low Back Pain. Spine 5: 78-86


Up to top of page