top of page
  • Michele Curatolo

Epidural steroids: Long-term improvement, but ...

Epidural steroids for the treatment of radicular pain can be injected with an interlaminar or transforaminal approach, close to the nerve root, dorsal root ganglion, and spinal  nerve.
Nerves and needle trajectories for interlaminar and transforaminal epidural steroid injections.

Injection of steroids in the epidural space is one of the most common procedures to treat leg pain, with or without back pain, caused by a nerve problem in the spine. Steroids are potent anti-inflammatory medications. The rationale for their use is the assumption that inflammation of the nerves that exit the spinal cord (see picture above) causes the pain, and therefore injecting the steroid at the site of inflammation will lead to pain relief. The figure above shows the two possible approaches: interlaminar (through the space between two adjacent spine segments), and transforaminal (through the space where the nerve exits the spine).

Do epidural injections provide long-term pain relief?

In a recently published study [1], we have addressed the question of long-term pain relief in older adults with a new episode of back pain, leg pain, or both. We have used our BOLD registry, a prospective study performed in three different health care systems in the USA, with a two-year follow-up. The study was focused on patients ≥65 years old. This is of particular importance because of the high prevalence of pain in older adults and because clinical trials and systematic reviews do not typically focus on older patient populations. The great value of this study is the prospective collection of multiple outcomes for two years in over 5,000 patients, including physical function, psychosocial function, medication use, and several other outcomes.

Result: Less leg and back pain, less disability, and better quality of life

The figure below shows one of these outcomes, leg pain. The blue and orange bars represent patients of the BOLD registry who did and did not undergo epidural steroid injections, respectively. The former group had more pain before the injection, which then decreased substantially. The improvement was sustained after two years. Back pain, disability and quality of life had a similar course.

Leg pain in patients who did and did not undergo epidural steroid injections.
Curatolo et al. Eur J Pain. 2022 doi: 10.1002/ejp.1975. PMID: 35604636.

... but not because of the epidural steroid injection

Patients who received and did not receive an injection in the figure above are not necessarily comparable. In order to build a control group to which patients who underwent an epidural injection can be compared, we used a statistical approach called "propensity score matching". This consists in creating a control group that did not receive an epidural injection, but had the similar characteristics as those who underwent the injection, such as sex, pain intensity, presence of depression, use of opioids, etc. The results are shown in the figure below. In contrast to the comparison shown in figure above, the matched analysis revealed lack of difference in leg pain between those who received and did not receive an epidural injection. The same was observed for back pain, disability, and quality of life.

Leg pain in patients who did and did not undergo epidural steroid injections, after propensity-supra matching.
Curatolo et al. Eur J Pain. 2022 doi: 10.1002/ejp.1975. PMID: 35604636.
Overall, the study shows that patients who underwent an epidural injection had a decrease in pain, improvement in physical function and quality of life during two years. However, these improvements were unlikely the result of the epidural injection.

If not because of the epidural injection, why did patients improve?

One possible reason is the so-called "regression to the mean". Some patients experience alternating periods of worsening and improvement, and are more likely to have an injection during a "bad" phase. The pain would decrease also without the injection as the result of spontaneous fluctuations.

For patients with a new episode of pain, a spontaneous healing may happen during the same time an injection is offered ("natural history"), and one would erroneously attribute the improvement to the injection.

Another possible reason for improvement is the "placebo effect": patient's expectation of pain relief with the epidural injection leads to an improvement, even if the injection itself is not effective. This is a physiological phenomenon that involves different mechanisms, such as the release of substances that produce pain relief in the brain when a treatment is expected to help (learn more here). In order to verify this explanation, we would need a study with a placebo group (patients having a procedure involving the insertion of the needle, without injection of any medication or with injection of an inactive solution). Such studies with a long-term follow-up are hardly feasible, for logistical and ethical reasons.

Should then we offer epidural steroid injections?

In our study, we did not evaluate the effect of the steroids before 6 months. Several studies have shown short-term benefits of epidural injections, up to few months. In certain circumstances, if pain is very high and patients are disabled, short-term improvements are valuable.

While short-term benefits may be a goal, performing or not an epidural injection is unlikely to affect the long-term course. Most patients will improve in the long-term even if no epidural injection is performed. Those who do not improve in the long-term would have unlikely improved even if they had received an epidural. This information may help patients and health care professionals in their decision whether to proceed with an epidural steroid injection or not.


[1] Curatolo M, Rundell SD, Gold LS, Suri P, Friedly JL, Nedeljkovic SS, Deyo RA, Turner JA, Bresnahan BW, Avins AL, Kessler L, Heagerty PJ, Jarvik JG. Long-term effectiveness of epidural steroid injections after new episodes of low back pain in older adults. Eur J Pain. 2022 May 23. doi: 10.1002/ejp.1975. Epub ahead of print. PMID: 35604636.

811 views0 comments


bottom of page