Vertebroplasty: The Myth of the Most Effective Medical Intervention
For the past two years, the Lown Institute has published reports on the rates of certain low-value spinal surgery procedures in individual hospitals and states. Among the procedures assessed was percutaneous vertebral cement augmentation (VCA), such as vertebroplasty, for suspected pain from fragility vertebral compression fractures (VCF) due to osteoporosis.
This procedure involves an injection of medical cement into the spine. This is intended to potentially stabilize the compression fracture of the bone and possibly provide pain relief. While VCA was purported to show great success in early reports, subsequent (much more robust) studies have shown little or no benefit for patients compared to placebo controlled sham injections. Yet, this procedure continues to be delivered to tens of thousands of older adults across the U.S. each year, wasting millions and putting people at unnecessary risk of harm.
After last year’s Lown Institute report, a group of radiologists led by Drs. Glade Roper and Douglas Beall railed against the report on LinkedIn, arguing that “vertebral augmentation is the most effective medical intervention that exists” and “it would be medically inappropriate not to offer osteoporotic patients who suffer compression fractures.” After reviewing the studies they cite, I couldn’t disagree with them more. Here’s why.
Early “success” of vertebroplasty
Vertebroplasty is a particularly interesting procedure to consider. An incidental finding of a VCF is very common in older adults, 20-35% by CT. At the same time, back pain is also extremely common in the same population—with or without a fracture. This means there is a lot of unrelated overlap and uncertainty in clinical practice. Despite a high degree of clinical expertise it is difficult to distinguish fragility VCF pain from non-specific, postural, arthritic or overuse back pain.
This naturally sets up a “perfect storm” for indication creep and contracting value.
When VCA was introduced, the purported outcomes were absolutely astounding. Advocates of the procedure, frequently with clear financial interests, have enthusiastically and doggedly accepted the results of these poorly designed and executed early studies while blinded, randomized controls trials have been dismissed or distorted with contumely.
Martin et al. reported that “[v]ertebroplasty is very efficient for treatment of pain” with an 80% “success” rate (success was undefined), one that is astonishingly high for any back pain procedure and has never reproduced in a controlled trial. Still, this report continues to be cited 20 years later.
In another very frequently cited study by Deramond et al (1998), the authors reported on a series of patients with a compression fracture on x-ray and severe focal back pain for 3-4 weeks (actually the authors admit, as little as one week…) treated with cement injection:
“Immediate results were excellent in more than 90% of the cases, with complete relief of symptoms and patients capable of standing up and walking 24 hours after the procedure. The analgesic effect was prolonged, as proved by a long-term follow-up.”
Ninety percent immediate, complete pain relief! Another, response virtually unheard of in the treatment of any serious spinal condition and apparently made by the authors (radiologists) without apparent irony. Even more impressively, the authors claimed they “proved” this “complete relief” lasted “long-term.” And how was this “proved”? Why, naturally, with follow-up by a radiologist for the “long-term”, which for those radiologists — is at least “one month”. Yes, this foundational study followed patients for one whole month. No pain scales, no standardized forms, no independent assessor…in fact no details of follow-up at all. Since this was “proved” with the certainty any Euclidean theorem, this article has been cited more than 800 times, including as recently as this year. The vertebroplasty and kyphoplasty boosters continue to resurrect the ridiculously erroneous to keep the faith with their “irrational exuberance.”
Better evidence comes in, showing lack of benefit
With such “proven efficacy” in mind, and very little more supportive evidence, vertebroplasty and its variant, kyphoplasty, have become both big business and business as usual. More miraculous case-series, pseudo-controlled trials with unvalidated or poorly matched open controls continued to inflate efficacy claims. They are all resting on a very slender reed. Multiple randomized, blinded, sham controlled clinical trials later repeatedly show no or minimal advantage over the very substantial and reproducible placebo, nocebo, or natural history effects. But better science has proven to little avail in curtailing overuse.
Higher level evidence from multiple Level 1 sham controlled studies which have categorically refuted early fantasies, has had little effect on the faithful to the gods of cement. Instead, they have further exaggerated the rhetoric—so much so that now vertebral augmentation is not only of high value, but it also ranks with the very highest value of all in medicine.
To wit, radiologists Glade Roper, MD and Douglas Beall, MD have informed us that unbeknownst to much of the world including the Lown Institute, “vertebral augmentation is one of the most effective medical interventions that exists to not only prevent pain and suffering but prevent premature death”. Well, we are impressed. Obviously, we thank Drs. Roper and Beall for correcting the generations-long failure to recognize that such VCA procedures belong amongst the greatest “medical interventions” ever. VCA we now learn is in the same league as water-purification, sanitary housing, vaccinations, general anesthesia, sterile surgical techniques, perinatal care, antibiotics, blood transfusions, rigid fracture fixation, large joint arthroplasty, insulin, antiretroviral medications, and hundreds of others. While penicillin alone indisputably decreased a world of suffering and saved hundreds of millions of lives, all these other medical efforts must share a pedestal with those who fearlessly inject cement into a vertebra to miraculous and unreproducible effect.
Mirabile dictu, indeed.
This absurd position is now firmly entrenched, thanks to the overlapping fields of fire from medical “influencers”, industry PR consultants, sacred fiduciary responsibilities, market analysts, infomercial publications, and les plus en vogue “content creators”1. The practice and business of vertebroplasty is now a juggernaut. Its boosters often have lucrative and byzantine financial ties to the device manufactures whose products are used in tens of thousands of percutaneous cement augmentation surgeries per year. The market size for the device manufacturers alone is nearly one billion dollars and projected to grow even more in the next few years.
More recent trials have truly unimpressive findings
Drs. Roper and Beall cite the VERTOS and VAPOUR studies as evidence that that VCA is “one of the most effective interventions” with the claim:
“…double-blinded controlled trials have demonstrated a benefit in symptom relief to treatment of compression fractures with vertebral augmentation, such as the VAPOUR and VERTOS trials.”
The VERTOS and VAPOUR studies were all funded by the manufactures of vertebroplasty devices and multiple authors, including some senior authors, had financial relationships with vertebral augmentation device manufactures. We know from research that such spine studies systematically favor the reporting of positive findings when they are industry-sponsored by an odds ratio of 3:1 to 4:1. So the negative and tepid findings in these studies are all the more startling.
We will quote the VERTOS IV authors’ conclusions verbatim:
“According to this study vertebroplasty in the early phase of a painful osteoporotic vertebral compression fracture does not provide important pain relief compared with a sham procedure.”
“Percutaneous vertebroplasty did not result in statistically significantly greater pain relief than a sham procedure during 12 months’ follow-up among patients with acute osteoporotic vertebral compression fractures.”
Perhaps not quite the same impact as the use of general anesthesia.
But there is more. The VERTOS V authors listed as its primary outcomes measures the VAS pain scores at 1 day, 1 week and 3, 6, and 12 months. At no time point was there a better outcome for the vertebroplasty group greater than the minimally clinically important difference (MCID) compared to a sham procedure. Not one. The same findings held for the secondary functional outcomes. There were only minor differences between groups, sometimes in favor of the control group. Figures provided show overlapping 95% Confidence Intervals at every follow-up point for both pain and function scores2. Analgesic usage was the same for both groups at every time point. At the end of 12 months both groups had moderately-severe self-reported pain and functional impairment.
For the faithful expecting anything like 90% complete long-term pain relief, these results should have been a shock. Instead minimal differences were celebrated as if this study had reproduced the 1955 Salk vaccine RCT. That trial actually did show a 90% rate of complete and lasting effectiveness of the vaccine vs. 0% for the placebo.
The VAPOUR study, 2016, was also industry sponsored, again by the CareFusion Corporation. The authors state the primary outcome measure was the NRS [Numerical Pain Score] pain score at 14 days after the intervention. The primary endpoint was the proportion of Numerical Pain Score (NPS) rating < 4 at 14 days after the intervention. The “principle” functional outcome measure was the reduction in RMDQ.
Again, the outcomes were unimpressive for this “most effective medical intervention.”
The results of the NPS and “principle” outcomes at 14 days show some differences between groups, but all inter-group differences are very small, and well below the minimal clinically important difference (MCID)3 for each outcome measure. There was no statistical difference in VAS or RMDQ outcomes at that timepoint between groups.
Post-hoc, subgroup and auxiliary analyses and cherry-picked subgroup analyses purported to show a better functional outcome in the vertebroplasty group at some timepoints. But post-hoc and auxiliary analyses are just that – auxiliary findings – they are most important when corroborating. Both arms in the VAPOUR study had a greater than 30% reduction in the RMDQ (the MCID) at 6 months. Both groups’ mean functional outcomes reflected persistent moderate disability. And in both arms, more than half the subjects required analgesic medication through one year follow-up (n.b. the control group actually started with more patients taking strong opioids).
In summary, the best “evidence” suggested by Drs. Roper and Beall shows one study with no benefit at all (VERTOS IV), another with no or minimal/ clinically unimportant differences (VERTOS V), and mixed primary pain outcomes no functional advantage in the VAPOUR study.
Vertebroplasty saves lives?
Unperturbed, Drs. Roper and Beall do pull one more rabbit out of their hat.
Apparently, unbeknownst to us all, vertebroplasty is one of the most effective life-saving interventions in the elderly (even if it has no or minimal functional or pain advantage). They proclaim, in their declarative sentence style, “for every 15 procedures done, one premature death is prevented.” Such is their categorical and unqualified opinion. And it is a categorical and unqualified absurdity.
In large dataset analyses, several groups have found that just being hospitalized with a vertebral compression fracture is associated with, among other admitting diagnoses, a much higher death rate in the next year as compared to those patients who enroll in RCTs of elective study procedures. In other words, patients in such observational studies are usually significantly sicker. Without adjusting for the dozens of confounding medical, social and economic factors that are known to predict life expectancy, what is one to make of the fact that those who received a vertebroplasty and who were, perhaps, healthy, wealthy or wisely-insured enough, lived longer? This is an unadjusted tertiary-level observation of an associated variable4. It is at best a hypothesis generating finding. To jump, as Drs. Roper and Beall do, to declare “for every 15 procedures done, one premature death is prevented” as established medical fact is frankly an embarrassment. A second year medical student would peremptorily fail their exam in “Methods and Design in Clinical Research” with such nonsense.
Owning a Gucci suit or Chanel handbag is undoubtedly associated with greater longevity, shall we intervene with a gift of a Gucci suit upon hospitalization? It may be cheaper than vertebroplasty. And, to our knowledge, no Chanel bag has embolized to the right pulmonary artery, nor thrombosed a paravertebral vein or even infected a disc. Gucci evening jackets are not reported to cause thermal paraplegia, nor do their trousers cause cardiac perforation and Gucci related emergency takeback surgeries are unknown.
So, where have all the miracles gone? Whither has flown the oft-cited 90% “immediate”, “complete” and “lasting pain relief” of yesteryear? Must we await a future VERTOS MMXLIII? Or may we say, at long last, sirs, the hype makes no sense in reality and this procedure has simply not worked out as advertised. As Dr. Beall eloquently stated, but from within his looking-glass world:
“We can and should do better with these topics. It may be understandable when we don’t know the information and repercussions but now that we do know, we should understand that inaccurate information like this hurts people.”
The actually accurate information does not reflect in any way that VCA is “one the most effective medical interventions” or even a high-value intervention.
Variation in vertebroplasty rates
It is, therefore, not surprising that the review of practice patterns by the Lown Institute reveals the schism between the faithful and infidels. The Lown Institute report revealed that patients admitted to hospital with a diagnosis of osteoporotic VCF were, in some hospitals, rarely treated with vertebroplasty procedures (< 0.5%), whereas others were found to treat the same condition 100 times more frequently. At a few hospitals an astonishing 50% or more of Medicare patients admitted with a VCF get a vertebroplasty.
The Lown report noted some of this disparity was geographically associated (e.g., in Arkansas the rate was over 18%, whereas in North Dakota the rate was 6%). But even in the very same geographic area and hospital type rates varied wildly. The Massachusetts General Hospital is three miles from Brigham and Women’s Hospital in Boston. The former was found to perform vertebroplasty five times more frequently than the latter for each patient with the same diagnosis. Fractures in Brookline possibly follow a different natural history.
This is not a game
Lastly, we need to address the following hyperbolical argument from Drs. Roper and Beall, who wrote:
“It would be medically inappropriate not to offer vertebral augmentation to osteoporotic patients who suffer compression fractures.”
Even cursory consideration would dictate that the claim that “any patient” (as they say) with osteoporosis and a VCF should be treated with vertebral augmentation is ludicrous. There are likely more than 1,500,000 fragility fractures of the spine in the United States each year. That’s three VCA a minute. When will anyone sleep?
Since most of these fractures are transiently or minimally symptomatic or completely asymptomatic, does Roper and Beall’s ethos require we recommend VCA to them all? Does it matter that back symptoms of VCF are difficult to separate from the common back pain in the elderly? Shall we endorse treating “any” such patient with vertebroplasty? Shall 20-30% of older persons have a vertebroplasty? While we may agree this would be a great business model, it sounds like very bad public policy.
Failing to establish a minimal clinically important advantage over a sham or placebo control group—in now five sham controlled RCTs—while carrying significant risks of harms is actually the very definition of a low-value intervention. We respectfully suggest that weak interpretations of weak studies may not be the best medicine has to offer, and that a constant drumbeat of fantasy and misinformation diminishes us all. If there is a subset, likely a small subset, who can benefit from vertebral augmentation, the studies thus far have not identified who those patients are. The situation is definitely worsened by hyperbolic rhetoric such as that of Drs. Roper and Beall’s latest missive – which advocates that any and possible every VCF must be injected. Dr. Beall, contrary to every controlled trial has set the standard for balderdashy with simultaneous self-promotion and misinformation when acting as both author, reviewer and, conveniently, his own editor. Here’s a sample:
“Vertebral augmentation…one of my favorite procedures…demonstrably life-saving …additional life expectancy will be 2 to 7 years…produces the best results of anything we do with an average pain score that decreased from 9/10 to 1.4 in our patients at first post-treatment visit…”
Unfortunately, this is a pernicious fantasy, not ever remotely true in this real world.
Equally unfortunately, such promotional material is everywhere: in hundreds of glossy adds, YouTube videos, infomercials, throw-away journals, sponsored commentary, Dr. Beall’s self-referential book of miracle and wonders and on and on. “Flooding the zone” is nothing new. Their self-regarding attached comments are only a sample. That sort of reckless soap-box vitriol impedes good care and stifles needed research by promoting shamelessly a boom-town practice ideology. It is dangerous.
This is not a game. The wholesale application of VCA to tens of thousands of patients, with little information of who exactly it may help, is harmful, wasteful and certainly by any definition low-value. Patients should be told as much. If patients, families and referring doctors knew that – the most carefully selected patients may at best –have a 1 or 1.5 point improvement in pain score and no functional improvement compared to a sham procedure, and then maybe only at three months after the VCA, how many would sign up for this boondoggle? Can the public finally be disabused of the miracle glow of discredited hype? If so, the public would be well served. And maybe, in saner times, someday we may focus on those who this procedure may actually help in a meaningful way.
Footnotes
- Google search, September 1, 2025, shows >700 un-vetted videos promoting vertebroplasty.
- Oddly the authors of the VERTOS V study choose a p value of significance at 0.05, despite having > 5 endpoint comparators. The alpha should appropriately be adjusted (by Bonferroni or other methods) to <0.01 or less. This is clearly visualized in the Figures with widely overlapping 95% CI bars. In any event, the magnitude of difference in means between groups indicate the effect size if there is one is minimal.
- There is no validated MCID for proportional reduction of NPS.
- “The present study certainly had some weaknesses: it was a retrospective study, it did not include any information about clinical decision-making as to why some patients received augmentation procedures whereas others did not, and it included only patients who were covered by Medicare.” Glaser J. Time to reconsider. J Bone Joint Surg Am. 2013 Oct 2;95(19):e1461-2. doi: 10.2106/JBJS.M.00762. PMID: 24088981.