Epidural Injection Didn't Work? Why and What to Do Next

Published March 7, 2026 · 6 min read

By Tauri Urbanik, Pain Science Researcher

The injection was supposed to fix it

Your doctor looked at the MRI. Pointed to a disc bulge or herniation pressing on a nerve. "We'll inject steroids right at that spot to reduce the inflammation." It made perfect sense. But as millions of people with normal MRIs and ongoing pain have discovered, imaging doesn't always tell the full story.

The injection cost $1,500 to $3,000. You lay on a table while a needle went into your spine under fluoroscopy. You waited for the steroid to kick in. Maybe you felt better for a few days while the local anesthetic was working. Then the pain came back.

Nothing went wrong with the procedure. The problem is what the procedure assumes.

Epidurals assume the MRI tells the whole story

Epidural steroid injections treat inflammation at a nerve root. They assume the structural finding on the MRI is compressing a nerve, creating inflammation, and causing your pain. Reduce the inflammation, reduce the pain.

But research has shown that the structural finding your injection targeted may not be causing your pain at all. A systematic review of 33 studies found that disc bulges appear in 50% of pain-free 40-year-olds (Brinjikji et al., AJNR, 2015↗). By age 80, 96% have disc degeneration with zero pain.

If the disc bulge doesn't cause pain in half the population that has one, injecting steroids at that disc may be treating an incidental finding.

What temporary relief actually means

Here is something worth reconsidering. If the epidural helped for a few days, most people think: "The injection worked but wore off." But what actually happened?

The local anesthetic numbed the area. Of course the pain stopped temporarily. That proves the nerve pathway is functioning, not that inflammation was the problem. Once the numbing wore off, the steroid was left to do its job. If the pain returned, the steroid had nothing meaningful to treat.

If the epidural didn't help at all, that is even more telling. The nerve root your doctor targeted may not be involved in your pain at all. The same pattern shows up when nerve blocks don't work either.

A failed epidural is actually useful information

Reframe the failure. If the injection didn't provide lasting relief, your pain is probably not coming from where the needle went. The MRI finding may be incidental. The inflammation may not be the driver.

This points toward neuroplastic pain. A learned pattern in the brain that generates real pain signals independently of tissue damage (Woolf CJ, PAIN, 2011↗).

Brain-based approaches like Pain Reprocessing Therapy work with this mechanism. In a randomized trial, 66% of chronic back pain patients became pain-free after just 4 weeks, with results lasting 5 years (Ashar et al., JAMA Psychiatry, 2022↗). No needles. No steroids. Just retraining the brain's pain response.

Frequently asked questions

Why didn't my epidural steroid injection work?

Epidurals target inflammation at a specific nerve root. But if your chronic pain is neuroplastic (brain-generated), there may be no significant inflammation at that site. The structural finding the injection targets may not be causing your pain. Research shows disc bulges appear in 50% of pain-free 40-year-olds.

Why did my epidural only work for a few days?

The epidural contains both a local anesthetic (short-term numbing) and a steroid (longer-term anti-inflammatory). If pain returned once the numbing wore off, the steroid had nothing meaningful to treat. The temporary relief proves the nerve pathway works, not that inflammation was the problem.

How many epidurals should I get before trying something else?

Most guidelines recommend no more than 3 per year. But if the first one didn't provide lasting relief, additional injections targeting the same site are unlikely to produce different results. The failure itself may be telling you the pain source isn't where the needle went.

What should I try after epidural injections failed?

Consider that your pain may be neuroplastic. Brain-based approaches like Pain Reprocessing Therapy have produced 66% pain-free rates for chronic back pain without any injections. A quick assessment can help determine if your pain patterns match the neuroplastic profile.

References

1. Brinjikji W, et al. Systematic literature review of imaging features of spinal degeneration in asymptomatic populations. AJNR. 2015;36(4):811-816.DOI: 10.3174/ajnr.A4173
2. Ashar YK, et al. Effect of Pain Reprocessing Therapy vs Placebo and Usual Care for Patients With Chronic Back Pain. JAMA Psychiatry. 2022;79(1):13-23.DOI: 10.1001/jamapsychiatry.2021.2669
3. Woolf CJ. Central sensitization: implications for the diagnosis and treatment of pain. Pain. 2011;152(3 Suppl):S2-S15.DOI: 10.1016/j.pain.2010.09.030