How Successful Are Epidural Injections for Spinal Stenosis?

Epidural steroid injections for spinal stenosis produce modest short-term relief for many patients, but the best available evidence suggests they work less impressively than their popularity implies. A major clinical practice guideline, drawing on high-quality evidence, now recommends against their use for lumbar spinal stenosis causing neurogenic claudication, and the largest randomized trial on the topic found that adding a steroid to the injection offered no meaningful advantage over injecting a numbing agent alone. That does not mean nobody benefits, but the picture is more complicated than “get a shot, feel better.”

What the Strongest Evidence Actually Shows

The most cited trial on this question was published in the New England Journal of Medicine. It randomized patients with lumbar spinal stenosis to receive either a glucocorticoid mixed with lidocaine or lidocaine by itself. At six weeks, there was no significant difference between the two groups in disability scores or leg pain intensity. A secondary analysis that separated patients by injection type, interlaminar versus transforaminal, found the same result: no significant difference either way.¹ That trial sent shockwaves through the pain-management world because it challenged the assumption that the steroid component was doing the heavy lifting.

Building on that finding, a North American clinical practice guideline reviewed the full body of evidence and concluded that epidural steroid injections should not be used for lumbar spinal stenosis patients with neurogenic claudication. The recommendation was based on what the authors classified as high-quality evidence, a strong rating that sets it apart from the weaker recommendations made against other treatments like gabapentin or calcitonin.²

This does not mean every professional society has reached the same conclusion. Pain-management physicians continue to use the injections widely, and some smaller studies do report benefit. But the direction of the strongest evidence is clear: the case for epidural steroids in spinal stenosis is weaker than for other spinal conditions like acute disc herniations, where the inflamed nerve root responds more predictably to a local anti-inflammatory.

Short-Term Relief Is Real, but It Fades

Even within a skeptical evidence landscape, many patients do report feeling better in the first few weeks after an injection. Several trials show significant improvements in pain scores and disability measures from baseline through about two months. One randomized trial found that both steroid and steroid-plus-calcitonin groups improved through the second, fourth, and eighth weeks after a caudal epidural injection.³ Another trial reported significant improvement in pain and disability scores at 3, 6, 12, and 24 weeks in the steroid injection group compared with conservative management.⁴

The catch is durability. A prospective study tracking patients over a full year found that while pain decreased significantly in the first few days, the benefit eroded over the following months. By one year, the number of patients with significant pain had climbed back up. The authors concluded that the effect of epidural steroids decreases meaningfully over the course of a year.⁵ So the realistic expectation is weeks to a few months of relief for those who respond, not a lasting fix.

Does the Injection Route Matter?

Three main approaches exist for delivering the injection into the epidural space, and the choice of route turns out to be one of the more consequential decisions in the process.

The transforaminal approach threads the needle through the bony opening where the nerve root exits the spine, delivering medication directly to the affected nerve. The interlaminar approach enters between vertebral arches, spreading medication more broadly. The caudal approach goes in through the tailbone, making it technically simpler but less precise.

A comparative study of all three approaches found that while all produced significant improvement in pain scores, disability, and stenosis-specific questionnaire scores, the transforaminal route came out ahead. Pain scores and functional outcomes were better from one month onward in the transforaminal group compared to both caudal and interlaminar groups.⁶ A separate study comparing transforaminal and caudal injections specifically reported that about 90% of patients in the transforaminal group showed pain relief at six months, compared with roughly 55% in the caudal group.⁷

The tradeoff is risk. A review of the three approaches noted that the transforaminal technique carries a higher chance of complications like nerve injury or vascular compromise precisely because it goes closer to delicate structures. The caudal approach, while typically the safest trajectory, delivers the least precise medication spread. The interlaminar approach sits in the middle on both precision and risk.⁸ Interlaminar injections also showed solid results in a separate comparison, where nearly 90% of patients achieved at least a 50% drop in their pain score at three months, compared with 75% for the caudal group.⁹

Does the Steroid Even Help, or Is It the Numbing Agent?

One of the most surprising findings in this field is that the steroid itself may not be contributing much. The NEJM trial already showed no benefit of steroid-plus-lidocaine over lidocaine alone. A systematic review and meta-analysis examining this exact question found minimal evidence that epidural steroid injections outperform lidocaine alone, regardless of injection route. The review did note fair short-term and long-term benefit from the combination of local anesthetic and steroid, but the steroid’s added value over anesthetic alone was negligible.¹⁰

A separate meta-analysis comparing lidocaine alone to lidocaine with steroids across seven studies confirmed the pattern. Both groups showed significant improvement from baseline through long-term follow-up, but meta-analysis found similar effectiveness for pain and function at three months and twelve months.¹¹ A commentary in a major orthopedic journal stated the conclusion plainly: epidural injections with glucocorticoid and lidocaine did not confer additional benefit compared with lidocaine alone.¹²

This raises an uncomfortable question: if the numbing agent alone does the job, what exactly is happening? One possibility is that the act of injecting fluid into the epidural space physically displaces inflamed tissue or interrupts pain signaling, and the steroid is along for the ride. Another is that lidocaine’s ability to calm nerve irritability has a longer-lasting therapeutic effect than its brief anesthetic window suggests. Researchers are still working this out, but for patients, the practical takeaway is that the steroid component may be less important than many assume.

Particulate Versus Nonparticulate Steroids

Among those who do receive steroid injections, there is a meaningful choice between two steroid formulations. Particulate steroids like methylprednisolone and triamcinolone contain tiny crystite-like particles that sit in the tissue and release medication slowly. Nonparticulate steroids like dexamethasone dissolve fully and have shorter-lived anti-inflammatory effects.¹³

A systematic review found no statistically significant difference in pain reduction or functional improvement between the two types for either cervical or lumbar injections. Given the lack of clear superiority for particulate steroids and the risk of rare but catastrophic complications, all of which have been reported with particulate preparations, the authors recommended nonparticulate steroids as first-line agents.¹⁴ Those catastrophic complications include strokes caused by particulate matter entering blood vessels and blocking blood flow to the spinal cord or brain, a risk specifically linked to the particulate formulation in cervical injections.¹⁵

A meta-analysis muddied the waters slightly, finding that particulate steroids produced a marginally greater change in pain scores, but nonparticulate steroids actually led to a larger proportion of patients achieving at least 50% pain relief.¹⁶ Given the safety advantage and roughly equivalent outcomes, the trend in practice is toward nonparticulate formulations, particularly for transforaminal injections where the needle sits close to blood vessels.

When Repeat Injections Make Sense and When They Do Not

Many patients receive a series of epidural injections, often two or three spaced weeks apart. The evidence on repeats tells a clear story: if the first round does not work within about six weeks, more injections of either type are unlikely to help. A randomized trial following patients for twelve months found that repeated injections offered no additional long-term benefit when the initial injections failed to improve pain in the first six weeks. For patients who did respond early, the improvements held at twelve months, but the additional injections did not appear to be the reason. The trajectories of pain and function after three weeks did not differ by whether patients received steroid or lidocaine alone.¹⁷

The practical rule: if your first injection gives you noticeable relief, a second or third may be reasonable to extend it. If the first one does little, adding more is unlikely to change the outcome and simply adds cost and exposure to side effects.

Can Injections Delay or Prevent Surgery?

For many patients, the real question is not whether the injection will cure them but whether it can buy enough time to avoid or postpone decompression surgery. A retrospective study found that about 58% of patients who received epidural injections achieved enough pain relief to avoid surgery entirely.¹⁸ That number needs context: it comes from a retrospective design, which is less rigorous than a randomized trial, and “avoiding surgery” does not necessarily mean the stenosis stopped progressing. Some of those patients may simply have tolerated their symptoms or found other ways to cope.

Still, even skeptics of epidural steroid efficacy acknowledge that the injections can serve as a bridge. A patient who gets a few months of meaningful relief may use that window to strengthen their core through physical therapy, lose weight, or simply get through a period where surgery is not practical. Whether injections directly delay the need for surgery or just make the waiting more bearable is a distinction that may not matter much to the patient living through it.

Physical Therapy Compared With Injections

A question many patients face is whether to try injections or physical therapy first. The evidence suggests they perform similarly. A trial comparing epidural steroids with physical therapy in lumbar spinal stenosis patients found significant improvement in both groups, with no significant difference between them. Physical therapy patients showed somewhat faster improvement in pain and physical activity scores at the two-week mark.¹⁹

Combining the two does not clearly add disability-related benefit. A randomized trial that assigned patients to either injections alone or injections plus physical therapy found no significant difference in disability scores at any time point. Where the combination did shine was in quality-of-life measures: patients who received both had better emotional well-being, emotional role function, and perception of general health at ten weeks.²⁰ So adding physical therapy after an injection may not make your leg pain any better, but it could improve how you feel overall.

Who Is Most Likely to Respond

Predicting who will benefit from an epidural injection remains frustratingly imprecise. You might expect that the severity of stenosis on imaging would predict outcomes, but the evidence does not strongly support that. A systematic review found insufficient evidence to either support or refute the use of stenosis visible on MRI to predict injection outcomes.²¹

What does seem to matter is how compressed the nerve root is and how recently symptoms started. Two studies found that low-grade nerve root compression predicted a better response to transforaminal injections. A shorter duration of symptoms before treatment also correlated with better outcomes.²² Age, sex, body mass index, and injection level did not reliably predict success or failure in the available data.²³

The crude takeaway: if your symptoms are relatively new and your nerve root is not severely squashed, you have a better shot at meaningful relief. If you have had severe stenosis with chronic symptoms for years, the odds tilt against you.

Risks and Complications

Epidural steroid injections are generally considered safe, but “generally safe” is doing a lot of work in that sentence. Common minor side effects include temporary increases in pain at the injection site, headaches, and brief blood pressure elevation. A comprehensive review catalogued a wider range of reported complications including spinal fluid leaks, positional headaches in up to about a quarter of patients, adhesive arachnoiditis, urinary retention, allergic reactions, and intravascular injections in roughly 8 to 12% of procedures. Rare but devastating events include spinal hematomas, infections, neurological deficits, and stroke.²⁴

The most alarming complications have been linked to contaminated steroid preparations. The same review noted a well-documented outbreak in which contaminated steroid injections caused meningitis, stroke, paralysis, and 25 deaths, with over 14,000 patients exposed. That event in 2012 was an extreme case tied to a single compounding pharmacy, not typical clinical practice, but it underscored the stakes of quality control in compounding and sterilization.

For cervical epidural injections specifically, the risk profile is more concerning. Complications include epidural hematomas, infection, and strokes from particulate steroid matter entering vertebral arteries. These strokes can cause damage to the spinal cord, brainstem, and cerebellum.¹⁵

Diabetes and Blood Sugar Spikes

If you have diabetes, epidural steroid injections carry an additional consideration that your pain-management doctor may not emphasize enough. A study of 30 diabetic patients found that blood glucose rose by an average of about 126 points after an injection, from a mean of roughly 160 to about 286. The estimated half-life of this spike was about one day, meaning most patients returned to their normal glucose range within two days.²⁵

A separate study confirmed that blood glucose was significantly elevated one day after injection but had normalized by day two. Patients with an HbA1c above 7% and those on insulin experienced larger spikes.²⁶ A jump of 126 points is enough to cause symptoms and could be dangerous for someone with poorly controlled diabetes. If you have diabetes and are considering an epidural steroid injection, plan to monitor your blood sugar closely for the first 48 hours and discuss dose adjustments with your endocrinologist or primary care doctor beforehand.

The Cost Question

Epidural injections are not cheap, and most patients do not receive just one. A cost-effectiveness study found that the average cost per injection was about $600, with most patients receiving three injections over more than a year. The average quality-of-life gain was tiny. About 45% of patients did gain measurable quality-adjusted life years at a cost of roughly $62,000 per unit gained, a figure that sits near the upper boundary of what health economists consider acceptable. Meanwhile, 40% of patients actually lost quality of life, suggesting they went through the procedure, bore the costs, and came out worse.²⁷

A separate comparison of injections versus medical management alone for radiculopathy found no significant difference in total costs or cost-effectiveness ratios. Medical management alone came in slightly cheaper at both three and six months, though the differences were not statistically significant.²⁸ The economic picture, like the clinical one, does not make a strong case for injections as a default first-line treatment.

Why Technique Variability Matters

One underappreciated factor is how much the procedure varies from clinic to clinic. A survey of physicians who perform lumbar interlaminar epidural steroid injections found significant variability in technique, despite the existence of consensus safety guidelines. Standardization remains limited.²⁹ That matters because the volume injected, the steroid used, the specific level targeted, and whether the procedure is done under fluoroscopic or ultrasound guidance all vary. One comparison of ultrasound-guided versus fluoroscopy-guided caudal injections found similar outcomes in pain relief, function, and patient satisfaction between the two imaging methods.³⁰ But the broader variability in practice means your experience with epidural injections might differ substantially depending on who performs them and where. Asking your provider which approach they plan to use, what steroid formulation they favor, and whether they use image guidance are all reasonable questions.

Cervical Stenosis Injections

Most of the discussion around epidural steroid injections for stenosis focuses on the lumbar spine, but stenosis also occurs in the neck. A prospective study of cervical transforaminal epidural steroid injections found that about 58% of patients achieved at least a 50% reduction in arm pain at one month, rising to roughly 72% at three months. At twelve months, about 65% still had clinically meaningful arm pain improvement, and about 71% showed meaningful improvement in neck-related disability. About 18% of the study participants had gone on to surgery by the twelve-month mark.³¹ These numbers look more encouraging than much of the lumbar stenosis data, though the cervical study was smaller and lacked a control group receiving anesthetic alone. Cervical injections also carry higher-stakes risks, as noted earlier, because of the proximity to vertebral arteries and the spinal cord itself.