Respectful Insolence

If there’s one theme that’s run through this blog since the very beginning, it’s that the best medical care should be based on the best science. In other words, I like to think of myself as being far more for science- and evidence-based medicine, than I am against against so-called “complementary and alternative medicine” (CAM).

Unfortunately, even though the proportion of medical therapies not based on science is far lower than CAM advocates would like you to believe, there are still more treatments in “conventional” medicine that are insufficiently based on science or that have never been validated by proper randomized clinical trials. This is true for some because there are simply too few patients with a given disease; i.e., the disease is rare. Indeed, for some diseases, there will never be a definitive trial because they are just too uncommon. For others, it’s because of what I like to call medical fads, whereby a treatment appears effective anecdotally or in small uncontrolled trials and, due to the bandwagon effect, becomes widely adopted. Sometimes there is a financial incentive for such treatments to persist; sometimes it’s habit. Indeed, there’s an old saying that, for a treatment truly to disappear, the older generation of physicians has to retire or die off.

That is why I consider it worthwhile to write about a treatment that appears to be on the way to disappearing. At least, I hope that’s what’s going on. It’s also a cautionary tale about how the very same sorts of factors, such as placebo effects, reliance on anecdotal evidence, and regression to the mean, can bedevil those of us dedicated to science- and evidence-based medicine just as much as it does the investigation of CAM. Moreover, it’s a story that I first encountered four years ago and wrote about at that time. As such, writing a followup on it is one of the advantages of having been blogging for nearly five years now. It’s one of those times when blogging is very satisfying, because it allows me to follow the development of of a story over years and comment on developments as they appear.

This latest wrinkle should serve as a warning to those of us who might feel a bit too smug about just how dedicated to SBM modern medicine is. What will physicians do with the most recent information from very recently reported clinical trials that clearly show a very favored and lucrative treatment does not work better than a placebo?

I’m talking vertebroplasty, baby:

Two new studies cast serious doubt on a widely used and expensive treatment for painful fractures in the spine.

The treatment, vertebroplasty, injects an acrylic cement into bones in the spinal column to ease the pain from cracks caused by osteoporosis, the bone-thinning disorder common in older people. Doctors began performing it in this country in the 1990s, patients swore by it — some reporting immediate relief from terrible pain — and it soon caught on, without any rigorous trials to determine whether it really worked.

The new studies are exactly the kind of research that health policy experts and President Obama have been calling for, to find out if the nation is spending its health care dollars wisely, on treatments that work. A bill passed by Congress this year provides $1.1 billion for such so-called comparative effectiveness research.

The studies of vertebroplasty, being published Thursday in The New England Journal of Medicine, found it no better than a placebo. But it remains to be seen whether the findings will change medical practice, because they defy the common wisdom and challenge a popular treatment that many patients and doctors consider the only hope for a very painful condition.

These are the two studies in question, and they were published in last week’s New England Journal of Medicine. Whenever a journal as prestigious and widely read as the NEJM publishes two studies of the same clinical question in the same issue, as companion studies, it’s trying to send a message. This time, the message is loud and clear:

  1. Kallmes DF, Comstock BA, Heagerty PJ, Turner JA, Wilson DJ, Diamond TH, Edwards R, Gray LA, Stout L, Owen S, Hollingworth W, Ghdoke B, Annesley-Williams DJ, Ralston SH, Jarvik JG (2009). A randomized trial of vertebroplasty for osteoporotic spinal fractures. New Engl. J. Med. 361:569-579.
  2. Buchbinder R, Osborne RH, Ebeling PR, Wark JD, Mitchell P, Wriedt C, Graves S, Staples MP, Murphy B (2009). A randomized trial of vertebroplasty for painful osteoporotic vertebral fractures. New Engl. J. Med. 361:557-568.

Before I get into the meat of the studies, let’s take a look at what vertebroplasty is. Vertebral fractures in patients with osteoporosis represent a very difficult problem to deal with. The reason is that they can be extremely painful and, worse, very difficult to get to heal using conventional methods. Indeed, the pain can be so severe that powerful narcotics are the only thing that can control it. Given that it can sometimes take months for such fractures to heal (if they heal at all), “conservative” treatment is not very satisfactory. Not surprisingly, a search for treatments that could decrease the disability, pain, and even death from these fractures was imperative. In the 1990s, a procedure known as percutaneous vertebroplasty, which involved the injection of polymethylmethacrylate directly into the vertebral compression fracture under the guidance of fluoroscopy, was first reported and popularized. Initial anecdotal reports suggested a very rapid and effective relief of pain after the procedure. The other aspect of vertebroplasty that led physicians to believe that it could work to relieve pain from compression fractures is that the technique had some degree of prior plausibility. The concept was that the injection of cement into the fracture would provide rapid stabilization of the fracture and therefore rapid relief of pain. It’s never been proven that that is the mechanism, but, according to the principles of prior plausibility, there was at least a reasonable hypothesis as to how vertebroplasty might stabilize fractures and relieve pain.

Now let’s take a look at the introduction of Buchbinder et al:

Observational studies suggest that there is an immediate and sustained reduction in pain after this procedure is performed,5 but data from high-quality randomized, controlled trials are lacking. The best currently available evidence for the efficacy of vertebroplasty comes from one randomized, open trial involving 34 patients and two quasi-experimental, open, controlled, before-after studies that compared vertebroplasty with conservative treatment. Although each study showed an early benefit of vertebroplasty, methodologic weaknesses cast doubt on the findings. In particular, the lack of blinding and the lack of a true sham control raise concern that the observed benefits reflected a placebo response, an effect that may be magnified with an invasive procedure.

Does this sound familiar? For long-time readers of SBM, it should. I’ve written about it before, as have others, in two contexts. First, consider acupuncture. Time and time again, I’ve pointed out how acupuncture studies tend to be positive for early, smaller, less rigorously controlled studies but in larger and better-controlled studies the observed treatment effect disappears. This sort of progression is very typical for interventions that don’t produce an effect that is measurably greater than that of a placebo. Four years ago, I described how vertebroplasty was in the phase where anecdotal evidence and small clinical studies suggested that it was useful. Indeed, I even referred to vertebroplasty as an example of just how hard it can be sometimes to avoid scientific delusions. Before the studies I’m about to discuss the evidence supporting vertebroplasty was actually very similar to the state of evidence for acupuncture in that the best evidence for the efficacy of vertebroplasty were two unblinded trials comparing vertebroplasty with medical management. Not surprisingly, they were positive trials showing a benefit for vertebroplasty over medical management. That so many physicians could convince themselves that an unproven treatment works on the basis of personal observation and small pilot studies simply shows how easy it is to persuade oneself to believe what one wants to believe, no matter how scientific one views oneself.

Cold, hard science has a way of cutting through all that and revealing that we, too, can fall for placebo medicine just as easily as any CAM practitioner.

The first blast of cold, hard science is a multi-institutional study (Kallmer et al) involving five centers in the United States, five centers in the United Kingdom, and one center in Australia, led by David F. Kallmes, M.D. at the Mayo Clinic and Jeffrey G. Jarvik, M.D., M.P.H. at the University of Washington. According to the methods, “sites were selected on the basis of having an established vertebroplasty practice for osteoporotic fractures, an enthusiastic local principal investigator, and an available research coordinator.” The study was designed as a randomized, controlled, study, in which the experimental group underwent standard vertebroplasty, while the control group underwent a sham procedure in which local anaesthesia was infused, and verbal and physical cues, such as pressure on the patient’s back, were given. To complete the sham, the methacrylate monomer normally used to make up the glue was opened to simulate the odor associated with mixing of the cement. However, the vertebroplasty needle, the needle was not placed and the cement was not infused into the frature. Patients with fractures due to bone metastases, bleeding disorders, and a few other criteria were excluded. Overall it was a sound design, but the investigators had trouble recruiting patients to the trial. Indeed, the NYT article from 2005 describes just this problem for Dr. Jarvik’s trial:

In 2002, a group of researchers received a federal grant for a clinical trial that would be the first to rigorously assess vertebroplasty. But their study is faltering.

Patients in severe pain have proved unwilling to enter such a trial, in which they might be randomly assigned to get a placebo, and their doctors have been reluctant to suggest it. In 18 months, the investigators have been able to persuade just three medical centers to recruit patients, and only three patients have enrolled.

Now the investigators are looking for centers overseas, but they agree that the study’s prospects are dim and that its failure would leave critical questions unanswered.

Once a belief has taken hold that a procedure works, it’s sometimes hard to persuade patients and physicians that a randomized study is ethical. Imagine how hard a sell this study would be to a patient at the apogee of pain from a compression fracture. That’s why there was a crossover design, in which patients, after one month, were allowed to crossover to the other group if they did not have adequate pain relief at one month.

In any case, this study was a resoundingly negative study. There were 131 patients enrolled (68 vertebroplasties and 63 simulated procedures), and the two groups were well matched. Both groups noted immediate and comparable improvement in disability and pain scores immediately after the intervention. At one month, there was no statistically significant difference between the control and vertebroplasty groups. The only hint of a possible effect is that there was a trend towards a higher rate of clinically meaningful reduction in pain in the vertebroplasty group, although, quite frankly, this result strikes me as a bit of the old trick of looking at multiple outcomes, trying to find one that comes out statistically significant. Let’s put it this way. This study is about as “positive” as any of the acupuncture studies we’ve dissected.

The second study (Buchbinder et al) was performed in Australia by a group led by Rachelle Buchbinder, Ph.D. It, too, was a multi-institutional study, with a control group that underwent vertebroplasty. The sham procedure was slightly different for this group, and, in my opinion, more rigorous. They underwent the same procedures as those in the vertebroplasty group up to the insertion of the 13-gauge needle, and to simulate vertebroplasty the vertebral body was gently tapped. As was done for Kallmes et al, the cement was prepared so that its smell permeated the room. Unlike Kallmes et al, there was no crossover was permitted in the design of this study. A total of 78 patients were enrolled and 71 completed the 6-month follow-up.

Buchbinder et al reported results that were just as unimpressive as those reported by Kallmes et al. In fact, they were more so, so much so that I will quote the abstract:

Vertebroplasty did not result in a significant advantage in any measured outcome at any time point. There were significant reductions in overall pain in both study groups at each follow-up assessment.

Can you say “placebo”? Sure, I knew you could.

Like Dr. Jarvik, Dr. Buchbinder crunched the data, looking for any outcome that could be considered even trending towards positive. She found none. This was about as negative a study as you can imagine, even more negative than most studies of homeopathy and acupuncture. This study found zero evidence that vertebroplasty is anything more than an elaborate placebo. Like acupuncture and homeopathy, come to think of it, the only difference being that there was some degree of prior scientific plausibility. Prior plausibility made vertebroplasty worth studying; however, these two trials provide emphatic evidence that it does not work, so much so that it surprised the authors, even though the very reason that the study had been done was because of nagging doubts that reported results seemed too good to be true:

Kallmes did the study because he always though the results reported for vertebroplasty seemed too good to be true–everyone got good results no matter how much cement was injected or what technique was used. And the mechanism through which vertebroplasty provided pain relief was a bit of a mystery. He figured if vertebroplasty was as good as promised, it would be easy to prove. He never expected results as stone cold negative as they were.

This is, of course, quite true. If vertebroplasty worked so well, it shouldn’t have taken much of a randomized trial to show it. If there’s one principle in medicine and clinical trials, it’s that dramatic treatment effects are much easier to demonstrate than weak ones. If vertebroplasty was as potent a treatment as advertised, these two trials should have easily shown its benefits. They did not. Indeed, in last week’s NYT article, there were a couple of more revealing quotes from the lead authors of these two studies:

“I’m going to be the most reviled radiologist on the planet,” said Dr. David F. Kallmes, the first author of one of the studies and a professor of radiology at the Mayo Clinic.

Indeed he is. He just provided very strong evidence that a favored cash cow among interventional radiologists is nothing more than an expensive placebo. So did Dr. Buchbinder:

“It does not work,” said Dr. Rachelle Buchbinder, a rheumatologist and epidemiologist at Monash University in Melbourne, Australia, and the leader of the Australian team. Dr. Buchbinder does not perform vertebroplasty and would “absolutely not” recommend it to patients, she said.

Dr. Kallmes, who helped develop vertebroplasty and has been performing it for 15 years, said his team was “shocked at the results.”

Shock is understandable. As had been pointed out in both papers, vertebroplasty had become the standard of care for the treatment of vertebral compression fractures. Medicare and insurance companies had, on the basis of unblinded studies, begun to reimburse for the procedure. Dr. Burbacher put it well:

Despite evidence that is acknowledged to be inadequate as a basis for justifying reimbursement, public institutions have recommended reimbursement for vertebroplasty. A recent position statement from various American radiologic and neurologic surgical societies also recommended funding the procedure.6 These endorsements have resulted in a dramatic increase in the number of vertebroplasties performed. For example, an examination of aggregate fee-for-service data from U.S. Medicare enrollees for the period from 2001 through 2005 showed that the rate of vertebroplasties performed during that time almost doubled, from 45.0 to 86.8 per 100,000 enrollees.14 There are also reports of repeat procedures for unrelieved pain at previously treated vertebral levels16 and of the prophylactic use of vertebroplasty in normal vertebrae that were deemed to be at high risk for fracture.

Not only is the short-term efficacy of vertebroplasty unproven, but there are also several uncontrolled studies suggesting that vertebroplasty may increase the risk of subsequent vertebral fractures, particularly in vertebrae that are adjacent to treated levels, sometimes after cement has leaked into the adjacent disk; controlled studies have shown conflicting results.

In other words, there’s money to be made. A lot of money. Boatloads of money. Combine the financial incentive with “personal anecdotal experience” that suggests to both patients and the physicians who do the procedure that vertebroplasty “works,” giving interventional radiologists the feeling that they’re “doing good while doing well,” and you can see why the procedure’s use has taken off.

But what about all the reports of dramatic pain relief from vertebroplasty? They’re testimonials, of course. I’m going to go back to the 2005 NYT article to show what I mean:

Dr. Jensen knows firsthand how powerful such stories can be. In the late 1990′s, when vertebroplasty was new and many doctors were looking askance at it, she gave a talk to a group of doctors in Chicago.

“I could tell by looking at the audience that no one believed me,” she said. When she finished, no one even asked questions.

Finally, a woman in back raised her hand. Her father, she told the group, had severe osteoporosis and had fractured a vertebra. The pain was so severe he needed morphine; that made him demented, landing him in a nursing home.

Then he had vertebroplasty. It had a real Lazarus effect, the woman said: the pain disappeared, the narcotics stopped, and her father could go home.

“That was all it took,” Dr. Jensen said. “Suddenly, people were asking questions. ‘How do we get started?’”

Even after these two studies, there are patients who will not be convinced, as demonstrated in last week’s NYT article:

One patient in the study, Jeanette Offenhauser, 88, said she was convinced that the cement had helped her severe back pain, even after hearing the results.

One thing that people often don’t realize and even physicians sometimes forget is that surgery and invasive procedures provide one of the most powerful placebo effects there is; so it is not surprising that a procedural intervention like vertebroplasty could appear effective at the anecdotal experience level. A lesson in history might be in order here. There was an operation that was popular in the late 1930s through the 1950s for treating angina due to coronary artery disease. It was called pericardial poudrage, and it involved opening the chest and sprinkling sterile talcum powder on the heart. The idea was that the inflammation would cause angiogenesis (the influx of blood vessels into the heart to revascularize it). We know today that the inflammatory reaction thus caused is far too minor to make up for the loss of blood flow from a major coronary artery. Still, patients reported marked improvement of their symptoms. It wasn’t until the 1950s that an actual randomized blinded trial (the patients could be blinded but the surgeons couldn’t) was performed. It found no difference between patients who had poudrage and those who just had their chests opened and closed. (Imagine trying to get that one through the IRB today.) In other words, poudrage was no different than sham surgery.

So it appears to be with vertebroplasty. Indeed, if I may be so bold, these studies should render vertebroplasty dead. To paraphrase what John Cleese would say, “E’s passed on! This procedure is no more! It has ceased to be! It’s expired and gone to meet its maker! It’s a stiff! Bereft of life, it rests in peace! If docs wouldn’t keep nailing it to the perch it’d be pushing up the daisies! Its metabolic processes are now ‘istory! It’ off the twig! ‘E’s kicked the bucket, It’s shuffled off ‘is mortal coil, run down the curtain and joined the bleedin’ choir invisibile!! THIS IS AN EX-PROCEDURE!!”

Sorry. I got a little carried away.

Even doctors are having a hard time accepting these results. Even Dr. Kallmes himself argues that patients who want the procedure should still be able to get it, but only under the auspices of a clinical trial. That’s not entirely unreasonable, but Dr. Burbacher disagrees and says that these studies should be enough to abandon vertebroplasty. In an accompanying editorial, Dr. James Weinstein of the Department of Orthopaedics, Dartmouth Medical School, argued:

President Barack Obama has called for more comparative-effectiveness research as part of the American Recovery and Reinvestment Act. Although clinical trials are an integral part of such research, from a safety and effectiveness standpoint, data from clinical trials combined with those from registries or other large longitudinal databases are necessary to provide the best evidence. Americans prize advances in technology. However, if in major medical challenges, such as osteoporotic vertebral compression fractures, the alternative is to pay the cost of perpetual uncertainty, we need to support the research necessary to provide sufficient efficacy and safety information for patients to make a truly informed choice. Although the trials by Kallmes et al. and Buchbinder et al. provide the best available scientific evidence for an informed choice, it remains to be seen whether there will be a paradigm shift in the treatment of vertebral compression fractures with vertebroplasty or similar procedures.

That is indeed the question, isn’t it? Here’s the challenge. Given these two rather definitive randomized studies showing in essence zero benefit for vertebroplasty over a placebo intervention, what will we as science-based physicians do? Will we, as we should, abandon a procedure against which the evidence has been accumulating, culminating in two large randomized studies that found no real benefit in terms of pain relief? Unlike homeopathy, for instance, vertebroplasty has a complication rate. It’s low, but it’s there. Harm can be done, and these studies suggest there is no benefit to make the risks worthwhile. Or will we behave as CAM advocates behave and refuse to believe because we know from our own experience that it “works”? Will we become acupunturists using fluoroscopy to guide much larger needles into the spine in order to redirect the flow of qi?

What will we do?

Dr. Weinstein, ironically enough, suggests what may well happen:

In an interview, Dr. Weinstein, who does not perform vertebroplasty, suggested that rather than abandoning the procedure, doctors could let patients decide for themselves, by telling them, “This is a treatment option no better than a placebo, but if you want to consider a placebo because you might benefit from it, you might want to know that.”

A $3,000 placebo? If given the choice between a pricey $3,000 placebo and a CAM placebo like acupuncture, which is likely to cost less than that and not have the potential for complications, you know what? I’d pick the acupuncture! If placebo medicine is going to be the wave of the future, at least with homeopathy, acupuncture, and reiki the potential for harm is less than sticking needles into vertebrae and injecting cement. Really, if allegedly science-based practitioners make this sort of argument, then why not just offer CAM to patients as a placebo and get in on the action?

Here’s my prediction. These studies will not be enough to change practice, at least not in the short term. I wish I could say otherwise, but I can’t. A procedure as well embedded into the standard of care and as profitable as vertebroplasty is won’t disappear overnight or without a fight. Less mercenary, vertebral compression fractures are a serious problem for which other treatment options (bed rest, pain killers, and back braces while the fractures heal) are both slow and not palatable to patients or physicians, who want immediate results. However, unlike CAM practitioners, eventually physicians will yield to the weight of the negative evidence. There may be a few more studies, and it’s even possible that there may be found a subgroup of patients who actually do benefit from vertebroplasty, but eventually the procedure will either be abandoned or scaled back to patients who might actually benefit from it. The process may be far messier than we would like. It may take far longer than we would like. It may even take a turnover to a new generation of physicians for the process to be complete. But, make no mistake, science will eventually win out.

If it doesn’t, then we are no better than the homeopaths.

Comments

  1. #1 Anthro
    August 10, 2009

    Years ago I hurt my back, had no insurance, went to the emergency room where I was barely looked at, given some pain killers and sent home–no tests of any kind. I couldn’t stand or sit in any position for more than a few seconds. I laid of the floor on my stomach and took the painkillers every three hours (instead of four) and cried. I slowly got better, very slowly. I had numbness on one side for months and forced myself to walk in spite of it. After a year or so, I could walk normally and had no more pain, but I still have some numbness on the side of my foot which I have learned to ignore and no doc I’ve seen since (now have ins.) has anything to say about it other than “you could try some physical therapy–IF your ins. will pay” (they won’t).

    I’m not saying I had this type of fracture, but I am saying that things tend to get better if you just rest and give it time. I hate the way I was treated by the clinic, but perhaps it was a gift in the long run?

    On another note, Dean Ornish is now the doctor in chief of Huff Post. He and Andrew Weill are featured in a banner headline. They seem to be confusing prevention with a certain level of CAM, as usual co-opting lifestyle changes into CAM. Weill has a very shady past that isn’t widely known and Ornish is kind of a crank (have you ever tried his “diet”–I mean 10% fat doesn’t even let you eat a walnut here and there!). Why can’t Huff Post get a regular, mainstream doc to speak for them?

  2. #2 Russell
    August 10, 2009

    Physicians purposely practicing wu are walking a more foul path than an accupuncturist or homeopath, because the former claim to be based on science.

  3. #3 James Sweet
    August 10, 2009

    Good article. I agree that the treatment will probably be (mostly) eventually abandoned, but probably not right away. I’m actually a little interested in the typical length of time that science takes to correct its own mistakes. Also how long it takes for conventional medicine to correct its own mistakes, because clearly those two things are not the same… In this case, the science is done self-correcting, but now how long will medicine take to catch up? Interesting.

    I watched part of SiCKO last night, and even though I agree with the basic premise that universal healthcare is an obvious priority and it’s a travesty that the US didn’t get that taken care of a decade or two ago, I was prepared for lots of misrepresentations and stupidity — and sure enough I got ‘em.

    The biggest thing that bugged me was this implication that getting rid of privatized health care will get rid of wonky coverage, where people get denied “experimental” treatments or ones that are too expensive, but get approved for other stuff that is less efficacious. I expect this to be somewhat better with more intensive government regulation, but you still are dealing with the same limited resources (which means you can’t give everyone everything they want), and the problem of the profit motive is replaced by the somewhat-lesser-but-still-confounding problem of lobbying. (The main draw of universal healthcare is the first word in that sentence, which is not only the right thing to do, but also attacks a whole bunch of other problems like preventative care, etc. But the idea that the government will do some miraculously better job at deciding what procedures are covered… well, I don’t believe that any more than I do the GOP lies that government will do a fantastically worse job. Should be a little better, not much.)

    The reason I bring this up is because vertebroplasty could end up being an example of the deleterious effect of lobbying on gov’t-run insurance. Imagine an advocacy group of patients who “believe” their pain was effectively treated with vertebroplasty, raising money to lobby congress to make sure that this treatment continues to be fully covered under some hypothetical gov’t healthcare plan. This is going to be the type of issue that the average person just shrugs about, and if they do have an opinion, it will probably be something along the lines of, “Well, if those people say it helps them, then insurance ought to pay for it…” So the vocal constituency will be the advocacy group, and you’ll have a hard time getting rid of it.

    The door swings both ways too. I pointed out to my wife last night that one thing that will likely suck about Obamacare is that, due to this whole lobbying effect, it probably won’t cover abortion. Even in the case of medical necessity, the definition of “necessity” will probably be absurdly narrow. Private healthcare is polluted by incompatible profit motives; gov’t healthcare will be polluted by ideology. It’s not only dishonest to ignore this; it’s unwise not to be ready for this when (knock on wood) universal healthcare finally gets setup in the US.

    Let me just be clear once again that I am not making an argument against universal and/or gov’t-run healthcare. My employer-provided insurance paid a larger percentage of my wife’s brief flirtation with acupuncture than they did for the birth of our first child. I expect to have the outstanding hospital bills cleared up before he is a year old, at least. Or if not, maybe I can start an acupuncture business….

    I think the free market is an incredible tool, in that corporations operating under a free market will do a ridiculously good job at whatever they are financially incentivized to do. In the case of health insurance, the financial incentive is to deny care, so… duh. Creating a financial incentive to deny care and then getting mad when a corporation denies care is kind of like if I drilled a hole in the middle of my new plasma TV and then got mad at the drill because the TV wouldn’t work. And then said that all drills were evil and started making up conspiracies about Big Drilla. Asking corporations to please be nicer and stop seeking so much profit is like asking a drill not to rotate so fast — and the only other answer is to bring in the gov’t.

  4. #4 James Sweet
    August 10, 2009

    Physicians purposely practicing wu are walking a more foul path than an accupuncturist or homeopath, because the former claim to be based on science.

    Enh, I agree with your conclusion but not your premise. In my experience, plenty of acupuncturists and homeopaths claim to be based on science.

    I think physicians purposely practicing a sham treatment are walking a more foul path, but IMO it’s because they have been imbued with implicit trust via things like medical degrees, etc.

  5. #5 Clare
    August 10, 2009

    While I agree that a procedure that cannot be shown to be effective should not be used (and certainly shouldn’t be paid for!) on the other hand, there’s going to be the problem that patients will want *something* particularly after hearing that there was a procedure that some people swore by even though it didn’t actually work. Imagine telling people that — sorry, we’re back to just morphine! — after this. Continuing to offer the procedure may for some simply be a matter of the money; for others, it will be the line of least resistance.

  6. #6 spudbeach
    August 10, 2009

    Wow! Thanks for the comprehensive review.

    In physics, I’ve heard it said that you should never believe just one article. Always wait for the confirmatory article, by a different group, before believing something. After all, mistakes are made, and most surprising results are false. (I even heard a professor say that Physical Review Letters, _the_ hot shit physics journal, is the journal of incorrect results, as the articles are short and rushed.)

    But this is two articles, by two different groups, both showing nil. The wise thing to do now would be to say that this treatment is dead (dead as a dead parrot is perhaps a bit much, but only a bit).

    So, back to basic research. I would love to see some post mortem results here — has anybody done a study of whether the glue is still in place on the autopsy? Whether there is any discernible effect of the glue? Some sort of analysis based on time since procedure would be nice. Further, if there is any subgroup that would be helped by this, we might search for results among the dead.

  7. #7 Ned
    August 10, 2009

    A closely related issue is the use of various medical procedures known to be efficacious for one condition to treat some other condition where the efficacy is unknown, or worse, it’s known to be nonexistent.

    Health care “horror” stories of sort are posted all the time. It’s usually somebody who is gravely ill and whose prospects are bleak – advanced metastatic disease, pancreatic cancer, that sort of thing – whose insurance company has balked at providing some unusual and much more expensive therapy.

    The response is invariably to excoriate the insurance company for denying this person their “chance at life”. But a literature search in many of these cases either turns up nothing to support the use of such a treatment, or worse, compelling evidence that the treatment would be a waste of time.

    Mind you, I’m not saying there aren’t plenty of examples where an insurance company denies a known effective treatment simply because they think they get away with it and save money. But some of the cases being held up as examples of this just don’t pass muster.

    The downside, if you want to call it that, of evidence based medicine is sometimes – oftentimes – the evidence says “no”, not the “yes” everyone really wants to hear.

  8. #8 Todd W.
    August 10, 2009

    It would be great to see more of these kinds of studies done for other surgical interventions. I seem to recall another procedure involving the heart (not the one Orac mentioned) that may not really work, though people swear by it (drug-eluting stents, maybe?).

    The problem with testing surgical procedures, though, is, as Orac mentioned, getting it approved by the IRB. Lot of messy ethics involved.

  9. #9 synapse
    August 10, 2009

    Is there any chance that simply being anesthetized for a while helps with the back pain?

  10. #10 Pablo
    August 10, 2009

    If placebo medicine is going to be the wave of the future, at least with homeopathy, acupuncture, and reiki the potential for harm is less than sticking needles into vertebrae and injecting cement.

  11. #11 Pablo
    August 10, 2009

    Sorry, I forget to close my blockquote. Ugh. Everything in that comment is mine except for the

    “If placebo medicine is going to be the wave of the future, at least with homeopathy, acupuncture, and reiki the potential for harm is less than sticking needles into vertebrae and injecting cement.”

  12. #12 laura
    August 10, 2009

    On a somewhat related note: as a cancer researcher, any thoughts about the use of the anti-osteoporosis drug zometa to prevent breast cancer bone mets?

  13. #13 Ramel
    August 10, 2009

    Could damaged vetebra be replaced in a similar way to a hip joint? And if not, why?

  14. #14 James Sweet
    August 10, 2009

    But a literature search in many of these cases either turns up nothing to support the use of such a treatment, or worse, compelling evidence that the treatment would be a waste of time.

    Mind you, I’m not saying there aren’t plenty of examples where an insurance company denies a known effective treatment simply because they think they get away with it and save money.

    Heh, this was also very much going through my mind when watching the first half of SiCKO last night. I kept wondering, with all these horror stories where people were denied treatment, how many of them might have been trying to get a highly experimental or a dubious treatment that wouldn’t have been paid for by any sort of universal healthcare program anyway?

    It seemed like a lot of the movie was the same sort of meaningless cherry-picking of nightmare scenarios that conservatives like to do when they are arguing against universal health care, e.g. pick an isolated example of someone who got fucked over by Canada’s system and act like that is indicative of the system as a whole. It’s not valid when fat lying conservative douchebags do it, and it’s not valid when fat lying liberal douchebags do it either.

  15. #15 MikeMa
    August 10, 2009

    Orac,
    As usual, thanks for the education. Love having ammunition like this when my woo-loving relatives come a-spouting! It’s not just the “cement” failure either, its the whole package from understanding the studies and methods to the related historical comparisons. All fascinating and almost all very accessible.

  16. #16 Bronze Dog
    August 10, 2009

    Could damaged vetebra be replaced in a similar way to a hip joint? And if not, why?

    You just ended up reminding me of an episode of Aeon Flux. Woman gets a spinal injury, indebted by medical bills, but she can move around thanks to these little containers of yellow liquid she puts right into a hole in the small of her back. She ends up removing them so that she can twist her upper half through a fence to talk with her lover on the other side of the border.

    My relatively uneducated guess would be no, since you’d risk damaging the spinal cord in the replacement process, and mankind hasn’t really gotten the hang of nerves, yet.

  17. #17 David
    August 10, 2009

    great article.

    Back pain, uncomplicated or not, is a field of unproven therapies that’s just a perfect setup for bad treatments tkaing hold. someday I’d like to take the time to compare the treatment effect size for placebo (or sham-treatment) groups across studies, versus the “impressiveness” of the intervention. There are studies with control groups of: doing nothing, handing out medical education leaflets, recommending physical therapy, taking placebo pills, getting sham accupuncture, massage, and now sham surgery.

    my hypothesis: the bigger the mask the witch doctor wears, the better the result.

    This partly explains why fad treatments exist (at least, for diseases with “soft endpoint”). The science-based physician dispenses moderately effective therapy but wears a small witch doctor mask. The quack wears a big mask and dispenses ineffective therapy. The net result to the patient isn’t so very different.

  18. #18 Tsu Dho Nimh
    August 10, 2009

    This study found zero evidence that vertebroplasty is anything more than an elaborate placebo.

    Like the noted “psychic surgery” and some other elaborate placebos, a convincing line of patter and some ceremonial processes can do wonders for you for a while.

  19. #19 Tsu Dho Nimh
    August 10, 2009

    What will we do?

    More to the point, what will insurance companies do? If it’s been convincingly shown to be ineffective, will they continue to pay doctors who do it?

  20. #20 Tom Conlon
    August 10, 2009

    You’re getting a little too fired up based on a relatively small study on a procedure that has been critically supported for over 20 years. The same speed to judgment you’re criticizing for the popularity of Vertebroplasty is what you’re exercising by calling for its abandonment.

    You’re overlooking a few points that bring this study into question in my mind:

    1. Placebo ?: The placebo resembles a medial branch block
    2. Possible selection bias: only 30% opted-in
    3. Too small of a group. Study did not meet it’s original goal of 250…why?
    4. Was each vertebroplasty carried out with the same standards? Anything from altitude (viscosity of the cement) to individual technique can effect the outcome.

    BOTTOM LINE? More study is needed!

  21. #21 James Sweet
    August 10, 2009

    my hypothesis: the bigger the mask the witch doctor wears, the better the result.

    I suspect much the same thing, but I’m not aware of any data to support it… It would definitely be interesting to know.

    I’ve often wondered if this is why sham acupuncture seems to do so well. Sticking needles in a person is one big-ass mask…

  22. #22 Kim
    August 10, 2009

    I’ve always found it interesting that vertebroplasty isn’t offered to young healthy people with traumatic compression fractures (at least, not the young healthy people I know who’ve made a habit of jumping off of cliffs and getting into car wrecks and such). If the procedure is so effective, why won’t younger patients benefit from it as well? Are the risks too high for a patient who is more likely to heal well on his own?

    In any case, that does suggest to me one avenue for studying the procedure: offering it to younger patients who wouldn’t receive it otherwise. They’re less likely to have expectations built up from what has been done for others in their social cohort, so it seems like there would be less reluctance to potentially end up in the control group.

  23. #23 bigjohn756
    August 10, 2009

    A good friend of mine went in today for his second vertebroplasty. He was in severe pain before the first one, and soon after he had that ‘fixed’ he was in pain again and they discovered a second one. I’m not sure he was ever pain free between procedures. I am anxious to see if the second procedure has any effect.

  24. #24 Hildy
    August 11, 2009

    We are better than the woo-meisters, because we don’t attempt to delude the patient and say ‘this will help you’. If you tell the patient that it is no better than placebo, and the patient accepts…

  25. #25 jempye
    August 11, 2009

    Talking about biased….. Kallmes study had 43% of control patients crossing over to vertebroplasty, and 12% the other way, highly significant, and a clear indication that the sham patients did not do so well. Study was designed to study at least dounble thye number ofd patien ts, which if it had done so, is likely to have rendered the pain results at 1 months significant in favor of vertebroplasty. The duration was only one month, too short to assess the real outcomes. The injection of numbing drugs is not placebo. And on and on…. Orac, please make sure that when your mother has an acute fracture that she sticks with braces and narcotics! You’ll be considered a caring and loving son.

  26. #26 marcia
    August 11, 2009

    The pain specialist

    http://www.youtube.com/watch?v=7Fp1HKJ1IrQ&eurl=http%3A%2F%2Fnews.google.com%2Fnews%3Fq%3Dvertebroplasty%26oe%3Dutf-8%26rls%3Dorg.mozilla%3Aen-GB%3Aofficial%26client%3Dfirefox-a%26um%3D1%26ie%3DUTF-8%26sa%3DN%26hl%3Den%26t&feature=player_embedded

    who undoubtedly follows evidenced-based medicine for other treatments, like implantable pain pumps, nerve stimulators, etc., looks like he just got whacked upside his head. There are a lot of docs out there, today, very nervous, and they may show up at a town hall as another shouting crank. (Maybe jempeye is one of them.)

    Oh, and let’s not forget this:

    “Percutaneous vertebroplasty strengthens a vertebra, but it also increases the stiffness of the segment. This altered stiffness of the vertebral body may again alter the distribution of forces to nearby vertebrae and thus increase the risk of fracture of these bodies.”

    Occurrence of New Vertebral Body Fracture after Percutaneous Vertebroplasty in Patients with Osteoporosis
    http://radiology.rsna.org/content/226/1/119.full

  27. #27 jillbryant
    August 11, 2009

    Interesting.

    1) I just want to clarify because I’ve heard people called anti-science here so, what is the difference between “avoiding scientific delusions” or “medical fads” and being anti-science. (I know what I think anti-science is but I think there is a different definition here.)

    2) Are chiros still what you call “woo” or has that changed since there has been a recent study to show they can be effective?

    3) Are there any discussions of other back surgeries that haven’t been proven to be effective? I believe while investigating this a doctor being interviewed on NPR mentioned there isn’t proof of effectiveness in some other back surgery(ies)

  28. #28 HCN
    August 11, 2009

    Jill Bryant, for more in depth discussions on chiropracty go here: http://www.sciencebasedmedicine.org/?cat=4

  29. #29 Jillbryant
    August 12, 2009

    Thank you. I do look at that blog. I think this study: http://www.ncbi.nlm.nih.gov/pubmed/15125860, that was quoted there is looking about as valid as ones doctors have relied on for more invasive treatments.

  30. #30 jempye
    August 12, 2009

    Marcia, I leave the town hall meetings for you. The CBS/Katie Couric item (you tube) has all the depth, objectivity and understanding that one would expect from the popular media, i.e. not much. The information you posted about new fractures is meaningless when considered in isolation. The natural incidence of new fractures in osteoporotic patients who experience a first fracture and who do not receive any systemic treatment is ~ 20% over the next 12 months. The debate as to whether vertebroplasty or kyphoplasty change that incidence is ongoing, and definitive proof will require a much larger study than the ones just published. The Buchbinder study, where 80% used biphosphonates, reports 3/35 vertebroplasty and 4/36 sham patients with new fractures at 6 months. This is about 10% in 6 months, in spite of the systemic treatment. It is helpful to analyze information in the appropriate context….

  31. #31 marcia
    August 13, 2009

    Yes jempeye, let’s examine some data in the appropriate context:

    Repeated and Multiple New Vertebral Compression Fractures After Percutaneous Transpedicular Vertebroplasty.
    http://www.ncbi.nlm.nih.gov/pubmed/19652633?ordinalpos=2&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum

    Oh, and this comment:

    “Orac, please make sure that when your mother has an acute fracture that she sticks with braces and narcotics! You’ll be considered a caring and loving son.”

    Unnecessary. And perhaps wrong. A walker (not sure braces are really necessary and they weaken surrounding musculature) significantly prevents new fractures. An intrathecal narcotic pump? Ask those who have one if they were happy when their children suggested it. You’ll be surprised at what they have to say.

    Katie Couric? Don’t like her? Then try this:
    http://www.youtube.com/watch?v=KR50TIDFwQc
    And, don’t tell me this guy will make money off the study. In fact, it looks like it’s just the opposite.

  32. #32 jempye
    August 13, 2009

    Ok Marcia, reading 101, or, go further than just the abstract: “Among all 852 patients, 141 patients (16.6%) experienced new VCFs during the follow-up period.” That period was 24 months. Interestingly, this rate is lower than the reported spontaneous rate without vertebroplasty. That there are more adjacent fractures is normal, as the preferred location for fractures is 1: the thoracolumbar junction; and 2: the apex of the thoracic curve. So, go look for another article to prove your point, this one doesn’t. There are so many articles of such varying quality that you should be able to find one that supports your argument. A walker as a treatment for painful osteoporotic fracture? Have you ever walked with one, or asked your 85 year old grandmother in pain to use one? Intrathecal narcotic pump??? And now you are going to tell me that these pumps are very effective and have no complications, or, are almost as good as placebo…. Sorry, I also did not buy land in Florida. I have heard Dr. Kallmes, or, as you call him, “this guy” speak more often. Yes, “this guy” benefits from this study. Monetary in that he got NIH grants that helped pay his or his staffs’ salaries, and non-monetary as he gains fame or, if you want, notoriety. The results of the study are non-conclusive, in large measure because they enrolled less than 50% of the planned number. The huge cross-over makes any longer follow-up pretty much meaningless. Stay tuned!

  33. #33 Orac
    August 13, 2009

    I have heard Dr. Kallmes, or, as you call him, “this guy” speak more often. Yes, “this guy” benefits from this study. Monetary in that he got NIH grants that helped pay his or his staffs’ salaries, and non-monetary as he gains fame or, if you want, notoriety.

    Poppycock. Balderdash. Bullshit.

    Dr. Kallmes is an interventional radiologist. He made a significant part of his living doing vertebroplasty for the last 15 years (as was pointed out in the quotes I cited). He used to believe that vertebroplasty works but thought that the reports before his were too good to be true and, quite rightly, wanted more rigorous data to support it. Also, take a look at his disclosures from the article:

    Dr. Kallmes reports receiving consulting fees from Zelos Therapeutics and grant support from ArthroCare, Stryker, Cardinal, and Cook and serving as an unpaid consultant to Bone Support…

    These are all companies that make either the cement or the needles and catheters used for vertebroplasty. I didn’t mention this in my post, but perhaps I should have.

    No, my conspiratorial friend who is fond of the reverse pharma shill gambit, if anything, Dr. Kallmes had a far greater incentive to show a positive result in his study. Showing a negative result arguably hurt him; it’s doubtful that these companies will be as eager to pay him as a consultant now that he’s done a study that shows their cash cow to be no better than a placebo.

  34. #34 marcia
    August 13, 2009

    Yes, jempeye, studies show walkers significantly prevent new falls. And, they can treat pain. How? Immobility leads to pain which leads to immobility. Surely, you know that. Putting some weight on a walker allows one to be far more mobile when in pain, thus reducing pain in the long run.

    Intrathecal pumps work at 100x less or greater the oral dosage of narcotics. This allows for more activity increasing the rate at which osteoblasts strenthen bones.

    Studies are now being performed to determine adjacent vertebrae changes in response to vertebroplasty. They can’t be ignored.

    http://www.ncbi.nlm.nih.gov/pubmed/17073569?log$=activity

    First, do no harm.

    And, you’ve got to be kidding me that the negative outcome in Mayo doesn’t trump the wealth and prestige from the grant and short-term notoriety.

    Nice try.

  35. #35 jempye
    August 13, 2009

    Language, Orac, language….. Dr. Kallmes will continue to perform vertebroplasties, continue to do studies, continue to receive grants and consulting fees, and continue to be happily and remuneratively employed at the Mayo Clinic, irrespective of the results of his study, and so it should be. None of you on this blog so far have produced any convincing arguments that refute the valid scientific comments regarding the study. The study was very well designed, and if executed as planned, would have been incredibly powerful, even if the sham procedure is not really a placebo. As it stands, having enrolled only 131 of 294 planned patients, with a clear trend favoring vertebroplasty for pain at 1 months (p = 0.06), with statitistically significantly more patients (43%) crossing over to vertebroplasty than to control (12%), with over 400 patients (almost 25%) excluded for non-disclosed reasons other than the ones identified in the protocol, with a too low minimum pain level required for inclusion (3/10), and a very easy to achieve and not clinically meaningful definition of success (30% improvement in pain, e.g. from 3/10 to 2/10) this study does not allow any generally applicable conclusions to be drawn. Statisticians that have commented on the study agree that it is very likely that the one month results would have been significantly in favor of vertebroplasty if only the intended number of patients had been enrolled.

  36. #36 jempye
    August 13, 2009

    Marcia, immobility leads to pain? So that is why doctors cast broken wrists and ankles, or use internal fixation, and than take away the heavy pain meds. They are all sadists! One wonders where you picked up this type of wisdom? Yes, activity strengthens bone, and that is done by our little friends the osteoblasts in concert with their cousins the osteoclasts. Great! However, if there is a little bit too much movement, the little friends can work all they want, the fracture will not heal. Yes, many studies are being done to look at new fractures, both in the lab and in the clinic. One day we will figure it out, but so far we don’t know enough. What we do know is that you should not choke on your Fosamax, because that may give you severe indigestion, and if you take it too long it may dissolve and drop your jaws.

  37. #37 marcia
    August 13, 2009

    Muscles can lose 30% of their size within 3 days of an injury followed by disuse. Muscle atrophy can easily lead to pain. There’s a strong relationship between chronic pain, somatic dysfunction and muscle atrophy as dysfunction and atrophy lead to reduced proprioceptive output from atrophied muscles and subsequent lack of nociceptor inhibition.

    Of course, one waits for some fracture healing before starting a movement program. C’mon, jempye.

    Now, I found this very interesting, small study of exercise vs. vertebroplasty, vs. exercise/vertebroplasty and second fractures at Mayo.

    Take a look:

    http://www.ncbi.nlm.nih.gov/pubmed/18174007?ordinalpos=20&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum

  38. #38 jempye
    August 13, 2009

    Now this is indeed interesting data allowing the formulation of a hypothesis for further research.

  39. #39 Mark Davidian
    September 3, 2009

    I am one of the interventional radiologists who, according to your blog, makes money off of vertebroplasties. I have probably done more than one thousand of them in my ten years of practice. I would like to enlighten some of the discussion. I have talked to several residents who were involved in the U.S. trial with Kallmes as the primary investigator. They had, as you correctly point out, a really difficult time enrolling patients. This is not surprising. Many of my patients cannot get out of a chair or bed without help due to the severe pain. I don’t think it is difficult to imagine them not wanting to enroll in a study with a sham arm. Which leads to the next observation. They had to change the enrollment criteria to allow entrance of patients with Visual Analog Pain scores of three or less (on a scale of 10). I would never consider doing a vertebroplasty on a patient with a pain score of 3/10. I can’t imagine this being done by any of my colleagues.
    I think what you are overstating in your blog is the definitive nature of the studies. The definitive study is forcing everyone who is referred for back pain in the U.S. with a diagnosis of vertebral compression fracture into a randomized, double-blind study with a placebo arm. I think with the low percentage of patients presenting to their service being enrolled in the study is a problem. To think that this study is definitive, I think, is wrong.
    One last point about Kallmes to correct one of the bloggers. He is a neurointerventionalist, and vertebroplasty is only a part of his practice. Most neurointerventionalists do primarily intracranial work (he has published extensively on anuerysm therapy). He is in academics, and the pressure to publish in academics is intense. He published a study in the New England Journal of Medicine. That is a very prestigious event and accomplishment. I don’t revile him, but I do wish that some of the shortcomings of the article would have had more space. All randomized studies are not created equal, and to say that these two studies kill the procedure I think is somewhat disingenuous. I know it makes great print for CBS, Couric, and here as well to discredit the procedure and those who perform it. It fulfills people’s image of some physicians doing things for monetary reasons. But I think you are jumping to a conclusion that may be popular, but unwarranted at this juncture.

The site is currently under maintenance and will be back shortly. New comments have been disabled during this time, please check back soon.