— “No more effective than sham procedures,” evidence review finds
by John Gever, Contributing Writer, MedPage Today
February 19, 2025

Just a week after an American Academy of Neurology (AAN) committee found “limited efficacy” for epidural steroid injections to treat chronic back pain, an international panel is going much further, calling for an end to these and a host of other common interventions.

Organized by The BMJ and comprising chronic pain specialists, methodologists, and patients with chronic back pain, the committee — headed by Jason W. Busse, DC, PhD, of McMaster University in Hamilton, Ontario — issued “strong recommendations” against the following procedures for chronic axial pain:

• Epidural steroid or anesthetic injections, alone or in combination
• Joint radiofrequency ablation, with or without spinal injections
• Joint-targeted steroid or anesthetic injections
• Intramuscular injections of local anesthetics, with or without steroids

And for chronic radicular pain, the committee recommended against:

• Epidural injections of anesthetics, steroids, or the combination
• Dorsal root ganglion radiofrequency therapy, with or without epidural injections

In each case, the committee based its recommendation on meta-analyses of blinded, controlled trials, which showed that these interventions did not outperform sham procedures.

Their report appeared Wednesday in The BMJ as part of its Rapid Recommendations series.

Accompanying it was an editorial by Jane C. Ballantyne, MD, of the University of Washington in Seattle, who firmly backed the group’s recommendations while noting that they didn’t address every possible way the procedures might actually be helpful.

“Would the conclusions of the linked meta-analysis have looked different if series of injections rather than single injections had been studied, or if the injections had only been provided in the context of comprehensive rehabilitation, or if the injections were confined to acute exacerbations of chronic back pain?” she wrote. “These are all questions that future research must answer.”

If these procedures are so clearly ineffective, why do they remain standard practice in most clinics?

Busse and colleagues, along with Ballantyne, pointed to the revenue the interventions produce. “In the U.S., the average cost for a single epidural steroid injection is more than $1,000, and can be as high as $5,000, and the average cost for radiofrequency ablation is approximately $6,000,” the committee wrote.

Ballantyne also cited the history of pain medicine and the training path that pain specialists now typically follow. Decades ago, after effective local anesthetics and other treatments were discovered that genuinely did relieve acute pain, clinicians naturally figured that they could be effective against chronic pain as well. “But it never really panned out,” Ballantyne wrote.

“Today,” she continued, “the dominance of procedural treatments in pain clinics is perpetuated because trainees, including non-anesthetists, want to acquire procedural skills and gravitate to the anesthesia-run programs that teach them.”

Still, it’s the case that many clinicians believe these treatments work because at least some of their patients say they do. The BMJ panel’s recommendations were based on a lack of difference between active and sham interventions — without considering the absolute degree of pain relief reported by patients receiving the sham treatments. Indeed, nowhere in either the group’s report or the editorial does the term “placebo effect” appear. Thus, some might question whether the committee is throwing the baby out with the bathwater.

Meanwhile, the AAN review, which focused on epidural steroid injections only, did find evidence that the shots were somewhat helpful even when tested against sham versions. “They may modestly reduce pain in some situations for up to 3 months and reduce disability for some people for up to 6 months or more,” said lead author Carmel Armon, MD, of Loma Linda University School of Medicine in California, in a statement.

Cynics may observe, too, that The BMJ panel leader is a chiropractor. But the committee did recommend other treatments with better evidence of effectiveness, and they did not include any chiropractic procedures. Instead they listed cognitive functional therapy, exercise therapy, and pain reprocessing therapy. These, the panel remarked, are “less well paying, and more time-consuming,” than one-off interventional procedures, thus creating “a perverse incentive” favoring the latter.

Methods and Findings

Busse and colleagues performed meta-analyses of 81 trials, with 7,977 participants in total. Studies focusing on chronic axial and radicular pain were analyzed separately.

The included trials were rarely very large, with median enrollment of 64 patients (interquartile range 45-110). Back pain had to have lasted at least 3 months to be considered chronic. Baseline pain was rated at a median of 6.8 on a 10-cm visual analogue scale. Most trials had funding from nonprofit or internal sources; only one reported industry funding.

Although all the group’s recommendations were labeled as “strong,” members weren’t always very confident in the strength of evidence. They acknowledged “low certainty” about, for example, intramuscular anesthetic injections and joint radiofrequency therapy for chronic axial pain, and about epidural injections of anesthetics or steroids for chronic radicular pain. A total of 13 interventions were studied, such that epidural steroids, epidural anesthetics, and their combination, for example, were examined separately; thus, there were not always very many trials or many patients to be analyzed. In some of the trials, they noted, patients were blinded to their treatments but providers were not.

Busse and colleagues devoted considerable space to harms associated with the interventions. While only about half of the included trials reported adverse events at all — and thus the evidence was relatively uncertain — the panel’s “consensus” was that “interventional procedures for chronic spine pain were costly and may be associated with a small risk of moderate harms.”

Some examples were an 8.6% rate of pain or stiffness lasting more than 48 hours, a 1.4% risk of dural puncture, and a 0.7% risk of deep infections. “Catastrophic harms” appear rare but because reports of them had appeared only in case studies and databases, the committee couldn’t estimate a rate.

About 2 weeks after this story was published, the American Academy of Pain Medicine (AAPM) issued a “newsflash” taking issue with The BMJ committee’s findings and conclusions. It was attributed to the AAPM’s president, president-elect, and immediate past president and read, in part:

“[W]e have significant methodological concerns about the erroneous conclusions raised in the recent British Medical Journal (BMJ) publication entitled: ‘Common interventional procedures for chronic non-cancer spine pain: a systematic review and network meta-analysis of randomized trials.’ This study evaluated all interventional spine procedures and arrived at an overbroad conclusion that interventional spine procedures are ineffective and should not be available to patients.

“Please stay tuned for AAPM’s detailed response to the challenges and pitfalls of these articles and next steps regarding these flawed publications. Our goal is to continue to promote team-based access to all effective treatments for our patients.”