Friday, 19 April 2013

Tennis elbow treatment: perception versus reality

For decades, surgeons have been reporting good results with surgery for tennis elbow. In a classic article from 1961, the late, great British surgeon RS Garden, reported that the results of surgery for tennis elbow were such that “no patient failed to benefit in some way from the operation”. Fifty years later in a review of 80 patients undergoing surgery for tennis elbow, 78 were reported to have improved. There are plenty of non-surgical treatments out there for tennis elbow (lateral epicondylitis) - all of them are reported as having good results, yet none of them are any better than placebo. Why then, did it take until now for a randomised trial to be done comparing real surgery with sham surgery?

The sham surgery trial
The study, of course, showed no advantage of real surgery over sham surgery. And, of course, patients in both groups get better. There was, however, one difference in outcome between the sham and real surgical patients; the surgical patients had more pain at two weeks.

Other treatments for tennis elbow
Patients improve after any treatment for tennis elbow, they just don’t improve because of the treatment. Take steroid injections, ultrasound (shock wave) therapy, orthotics and acupuncture – they are all perceived to ‘work’ (in that patients usually feel better afterwards), but they all fail when put up against a well-blinded placebo. (Review of corticosteroid injections here, Cochrane reviews of shock wave therapy here, orthotic devices here, and acupuncture here).

The classic article
The article by Garden from 1961 is worth a read, if only to see the standard of surgical research at that time. It is actually a comprehensive case series, it just contains so many logical and methodological flaws (that are largely hidden to the reader and the author). Most telling is Garden’s attitude: it was clear to him that the operation was successful, and in a sense he was right. It’s just that the patients didn’t get any better because of the surgery, and his study did not prove that they did.

Garden also said that surgery was usually not necessary because the results of steroid injections were so good. Yet at best, steroid injections give short-term benefit and then that’s it. I reckon if you didn’t mix them with local anaesthetic, you wouldn’t even get the short-term relief.

The bottom line
Nearly every operation being performed for subjective complaints (usually pain) today has not been tested against placebo and is performed for the very reason that Garden did his operation on tennis elbow: because the patients seem to get better afterwards (see my previous blog post for a more detailed explanation of the ‘wobbly tripod’ of evidence for surgery). 

I was going to say that this is a case of (the logical fallacy of) post hoc, ergo propter hoc on steroids, except that steroids don’t work either.


  1. An interesting study, but it should have been noted in your post that this is a pilot study. The pre-study power analysis indicates that 80 subjects would be needed, though admittedly the post-hoc power analysis of this pilot study seemed good. My statistics are pretty rusty, so I'm just wondering if the proper statistics were used for these measurements? Some other things to note with these subjects is their function level and exactly how much they utilize their elbows, i.e. population control. Would the same results occur with construction workers? Tennis players? Or is this just for a general population of patients?

    I think most orthopedic doctors I know nowadays tend not to use surgery as their first weapon of choice like they may have in the past. As a physical therapist, I see a ton of patients from orthopedic and neuro specialists who want desure conservative treatment first, followed by or in conjunction with steroid shots, epidurals, topical analgesics, and other forms of pain/anti-inflammatory controls before using surgery as a last option. Of course I may be a bit biased in what I see since I don't sit in the specialist's offices and watch every patient walking through.

    1. Thanks, I was waiting for some comments about the study. I was going to (and still might) add a small section at the bottom with details of the study, because it only included 22 patients. I guess my point about the study is that it showed no difference (in fact surgery was worse for pain) so the likelihood of surgery being better than placebo is low, even with a larger study, and certainly lower than the effectiveness that is seen in case series. Which reinforces my larger point, that so many treatments appear to be effective, but fail when tested against placebo.
      I don't have any figures on rates of surgery over time, but it is telling that people are still publishing large case series. Also, it is interesting to note that the article from 1961 stated that surgery was only a last line option and usually not required (because steroids were so effective!).
      Most of my colleagues know that surgery for tennis elbow is a last resort (same for plantar fasciitis) and may not be effective. But they still do it!

  2. Tennis elbow surgery is definitely last resort. Degenerative tears in all manner of tendons are part of aging and loss of repair capability of the body in parts such as rotator cuff, hip abductors, Achilles etc. One pilot trial for tennis elbow does not make all surgeries for tennis elbow less than placebo. Like rotator cuff repairs, some fail and some work in treating pain/ function of which the real reason is yet to be determined. Although, the surgery does not restore the muscle/ tendon unit back to its original state especially in chronic tears.
    The placebo effect is very powerful, granted that the surgeon and the patient both are convinced about the effect. How do you consent to placebo surgery as a real effect if the surgeon is convince it is placebo. How do patients suffering from chronic tennis elbow for 2 years, have an MRI that suggests a tear, and then told there is no cure, live with it?
    Compare that to the same patient, visits a surgeon, and be told it will work 50% of the time. Although no better than placebo, 50% chance of improvement is still better than no chance. Catch is both surgeon and patient has to be convinced that it is NOT placebo.
    I admire the push for surgeries better than placebo and EBM so after the pilot, do a QoL and DASH for these patients 5 years post randomisation for placebo versus surgery and see if any lasting (perceived or real) effect especially in those who initially surgery and placebo had worked.

    1. Thanks.
      Yes, this trial does not mean that other types of surgery for tennis elbow are not effective, but it does not allow us to assume that they are, which is what we tend to do. However, I would be willing to put money on the results of any blinded placebo surgery trial for tennis elbow showing no improvement.
      Follow them for 5 years? What is the natural history of tennis elbow? They will all be better in both groups.
      How do patients with pain live with it? That is not the point. People have been "living with" back pain and arthritic pain and lots of other things for thousands of years, often a lot better than they are now. The point is: does the treatment work? The presence of symptoms does not make a treatment effective.
      I think you are saying that if we think it works and convince the patient of that, then that is OK because the results are better? As long as there is no increase in harm (and as long as I am not paying for it), then go ahead, but then you can't criticise homeopaths and alternative medicine providers for doing exactly the same thing.
      You state that "The placebo effect is very powerful". I disagree. By definition, placebos cannot have an effect. See my comment from today on prolotherapy which addresses the absence of any real placebo effect:

    2. 5 years is a too long, I was making a point as to the natural progression or the condition. How would you know that they all get better? I have not encountered any evidence for this?
      "The presence of symptoms does not make a treatment effective" so as a counterpoint, the absence of symptoms make the treatment effective.
      What I am saying about placebo is that, it is a real effect. The effect may be conditioning, duplicity or by another means. Just not an effect that is intentionally physiologic, or pharmacologic. I agree with the definition of placebo but I disagree about the placebo effect.
      I cannot explain, why pain is felt and interpreted differently in people having the same injury/ condition/ pathology. Part of this is psychological rather than neurobiological - or at least, the neurobiology is not well established.
      Sure a lot of people live with pain - I rather live without it. I dont have a problem with alternate medicine providers. They are last resort - like tennis elbow surgery.

    3. Thanks,

      The natural history of tennis elbow is that it is usually self-limiting; it gets better. Whether they all get better or not is not the issue, it is whether they get better because of surgery or not.

      Whether or not we call it the placebo effect, your point is right: there are many reasons why people might get better, or not get better. It is complicated, so the only way we can tell that a treatment has any true, specific effect is by testing it against a control.

      Doctors use the perceived effect of ineffective treatments all the time, my problem is that they believe them to be effective and they are wrong in that belief. The doctors belief makes it more effective, true, but that is not what doctors are saying, they are saying the the improvement is due to whatever biologically plausible mechanism they have chosen to explain the improvement. That is not how I want to practice medicine, or spend the public's money.

      Your counterpoint is logically false. If the presence of A is not causally linked to B, then you cannot assume that the absence of A is linked to B. The presence of pain, of whatever magnitude (including zero), does not make an ineffective treatment effective. This is a common reaction from patients when I explain to them that surgery is unlikely to help their degenerative back, or worn out hip, or tennis elbow, or whatever. They then emphasise the severity of their symptoms, so I have to explain that the severity of the symptoms doesn't make the surgery any more likely to be effective.

    4. Placebos can't have a specific effect, by definition, but that doesn't mean that the entirety of what's called "the placebo effect" is regression to the mean, natural history, measurement error, etc.

      The perception of pain is what pain is, neurophysiologically. Pain doesn't exist apart from our perception of it: "pain" is a percept. Intuitively, it seems that pain must be caused by tissue injury or other objective causes ("cause" in the sense of an external process directly changing the activity of our perceptual apparatus in such a way that we perceive pain). And of course many, many injuries and diseases cause pain in such a way that the relationship between the tissue insult and the sensation of pain isn't problematic.

      But pain -- even (and somehow, most likely) intense and prolonged pain -- can and does develop in normal or at least fairly unremarkable tissues. The classic neurological example of this is migraine, in which headache has a neurophysiological basis (altered patterns of cortical activity, blood flow, neurotransmitter release, etc) that is episodic, fully reversible and not associated with tissue damage as the cause or result of the syndrome.

      Chronic somatic pain from the neck down is often diagnosed as musculoskeletal even in the absence of definitive or unusual pathology to explain it. For example, chronic back pain has been attributed to "disc bulges" at least since MRI became widely available. Many (most, by a certain age) people who have little or no back pain have disc bulges of comparable size and number to those found in patients with chronic disabling pain, so the relationship between what seems likely to be normal anatomic wear and tear and that pain syndrome is not clear.

      I don't know anything about tennis elbow, but from my clinical experience with other chronic somatic pain disorders it may be the case that the MRI tears being blamed for chronic intractable pain disorders are also found in paucisymptomatic patients. I just found one recent paper on PubMed that suggests that, while MRI can reliably show changes of tendinosis at the elbow, the severity of those changes *negatively* correlates with patient's reports of pain intensity ( This would suggest that patients with mild radiographic epicondylitis (assessed by blinded radiologists on MRI) are more likely to have severe pain than those with severe radiographic epicondylitis.

      To me, that's no surprise. A few obvious exceptions aside (cancer in the bones, etc), chronic intractable pain is *usually* a neurophysiological disorder not well explained by underlying tissue pathology. Similarly negative correlations have been shown for many types of chronic pain, most famously "whiplash" and concussion -- in both cases, more chronic headache and neck pain disorders are caused by low velocity impacts than high velocity.

      I would suspect that many patients with chronic intractable elbow pain diagnosed as epicondylitis (who are the only patients who will consider surgery -- if they weren't intractable, they wouldn't need the OR) have a secondary chronic pain syndrome (a neurophysiological pain disorder comparable in many ways to migraine) overlying the tissue injury that triggered their pain, amplifying it. You would expect a heterogenous population: 1. mild tissue damage plus chronic pain syndrome (who are enrolled in studies and eager for surgery), 2. severe tissue damage +/- pain syndrome (similar), 3. mild tissue damage without overlying pain disorder (who are at home with mild symptoms, not getting MRIs and meeting surgeons).

      (sorry this is so long, continued below)

  3. (part 2)...

    Benedetti and many others have shown on fMRI that when patients report a subjective benefit from placebo (pain relief), their brain states change in a way indistinguishable from those patients who have similar relief from real analgesics. That's a real change in neurophysiology; it's not imaginary just because it's subjective.

    So we would expect that placebos could cause subjective improvement (pain relief) without changing the underlying objective situation (tendinosis). The pain relief would be a real thing, with a verifiable neurophysiological basis, though at the same time it would be a psychological benefit (in the sense that all perceptual states are psychological, not in the sense that the pain "isn't real" or that the patient "is just depressed").

    Having said that, it's not the case that placebo analgesia for chronic pain in the elbow associated with tendinosis on MRI will cause that tissue injury to actually heal. Even if the patient feels the pain is gone, we will not see a significant difference in the tissue itself. And in that sense, the placebo is truly powerless.

    And yes, regression to the mean and natural history certainly make up a big part of the misnomer of the "placebo effect" in research -- but given the patient population of this study, with average duration of symptoms nearly 5 years before surgery/sham, one wouldn't expect too much of a benefit purely from those factors (you'd need another group randomized at week 0 but followed just for history without intervention to be sure).

    In terms of diagnosis and management of chronic pain, the mistakes we (doctors and patients alike) make are:
    1. to blame persisting severe pain entirely on a tissue insult, ignoring the pain experience that can and often does overlie that
    2. to expect that the pain can't improve unless "the underlying problem is fixed"
    3. to go to sometimes heroic or at least unjustified lengths to "fix" that problem
    and then, if the patient improves:
    4. to believe that the subjective improvement is entirely due to objective improvement (even though similar objective changes such as post-surgery do not reliably produce pain relief in many other cases)
    or if not...
    4a. see #1

    With this model it's easy to see why so many reasonable people can be convinced that placebos (including most CAM treatments and many mainstream medical / surgical procedures never proven better than placebo) are "really helping" them. They really do feel better, and not feeling well was the main thing wrong with them in the first place (the pain was more of a problem than the tendinosis, in other words), so they are not too badly off on balance, as long as the treatment wasn't too expensive, risky etc.

    1. Thanks Plasmon, excellent comments. On the whole, I agree with you, particularly about the mistakes that doctors and patients make about chronic pain. We tend to take the shortest reductionist pathway based on our knowledge.

      I think the most important point you make also touches on comments by Anon, above. I agree that regression, natural history etc. do not explain all of the response, and that relief may be due to other aspects of the therapeutic envelope. So your point then, is that if people feel better after the treatment, does it really matter if it was due to the expectations / therapeutic environment / placebo effect? Removing any added risk or cost from the argument, I think we really have a good question here. Historically, probably up until WWII, doctors have been practicing medicine like that, providing largely ineffective treatment with gravitas and confidence, and making people happy. Not all, but some, sure. Doctors are still doing it, as are other healers.

      Firstly, I accept your and Anon's argument that this is reasonable treatment, and that it continues to be reasonable treatment in our culture and in just about all other cultures. But allow me to list my objections. The question to answer then will be: Do my objections outweigh the benefit provided by these treatments?

      1. It is wrong to claim specific effectiveness where it does not exist.
      This is the problem with all the biological mechanisms we make up to explain why the patient got better. It is deceitful and falsely raises the effectiveness of medicine (along with doctor's power, wages etc.). This is being tackled by researchers who are providing placebo treatments without the pretense: by providing 'open-label' placebos (google Kaptchuk, placebos without deception).
      2. It is wrong to provide such treatment and criticise alternative providers for doing exactly the same thing.
      You may not do this, but many do, which amounts to a double standard.
      3. It is unscientific.
      You approach the problem very scientifically, wanting to understand it better. That is what we should be doing, not peddling pseudo-remedies and resting on our laurels. For example, simply educating the patient about some of the things you have raised can help them more than any placebo treatment. We should be applying the science as we know it, and seeking to expand our knowledge.
      (Objections 4 and 5 cover public cost and potential harms, but I am willing to waive those objections because they do not always apply).

      Will my way lead to more relief and less pain in the world? I honestly don't know, but at least we won't be kidding ourselves.

  4. I agree with you in general.

    Particularly agree that medicine has always made (mostly non-self-conscious) use of nonspecific treatment effects to help patients feel better, and until the advent of antibiotics, we had little better to offer. Benedetti has a book about the neurophysiology of the experience of receiving medical care -- The Patient's Brain.

    We do now have specific effective treatments to offer for many conditions, but our armamentarium may not be as robust as we like to think. Many of the medications we use for managing chronic/recurring subjective symptoms (including my territory, migraine prevention) may have their benefit in part because they produce side effects apparently unrelated to any treatment effect, which leads to unavoidable unblinding in RCTs versus placebo. Kirsch demonstrated this quite convincingly w/r/t SSRIs -- I think I commented while back on one of your other posts about his reanalysis of those data -- showing that the patients who improve in SSRI trials are those who believe they were on the study drug, regardless of which group they were actually in, and that the treatment effect over placebo is explained by the fact that a higher percentage of patients getting the SSRI correctly surmised they were in the treatment group while fewer getting placebos came to that same conclusion. I remember reading a similar debunking of topical capsaicin for chronic regional pain -- tested against an active placebo able to cause similar initial pain by different mechanisms, there was no benefit.

    I can easily imagine similar unblinding explaining the treatment above placebo for many of the migraine preventives I prescribe, which cause sedation / dry mouth / tingling etc. Botox injections are the new thing for chronic migraine, and the evidence is clear that they work better than placebo, but of course it's impossible for placebo IM saline injections to produce visible paralysis of the forehead. There would have been significant inadvertent unblinding as a confound (and even so, placebo groups improved considerably, but Botox groups had a definite statistically significant benefit above that). To avoid unblinding, you'd have to do a trial of Botox vs saline without injecting the forehead and in doses not likely to produce head drop or other noticeable side effects. Somehow I doubt Allergan will jump at the chance to fund that study.


  5. ... (part 2)

    The background for the recent debates about placebo is the "science based medicine" community (exemplified by the excellent blog of that name) fighting a turf war against various forms of CAM. Broadly speaking, SBM is dead set against there being any physical manifestations of placebo (their thinking is hugely muddled on whether changes in neurophysiology corresponding to subjective symptoms are "real" benefits), while CAM likes to promote "mind over matter" as an all-purpose hand-waving explanation for the supposed therapeutic benefits of various forms of consensual delusion.

    And I'm on the side of the SBM community in that fight. The CAM people certainly aren't humble enough to admit that their handwaving about subluxations or chakras or toxins or whatever are no more than Voltaire's adage of "amusing the patient". They make unsupported claims about joints healed, tumours regressed, brain damage healed, etc -- stuff that sorely needs debunking.

    But SBM (not the blog, the whole community of modern scientific medicine) has been guilty of overreach in my opinion. The laudable but (in my opinion) misguided effort to base all treatments on RCT evidence and treatment algorithms has made data collection more important than clinical judgment, and the parallel move toward making consent the centrepiece of the clinical encounter has pulled the curtain back on the Wizard of Oz effect that was historically the source of doctor's power and influence over the patient's wellbeing and in the community at large. I think it's no accident that the rise of this stricter, dispassionate version of modern science based medicine has coincided with the ascendance of CAM to fill the void -- being more than willing to engage patients' imaginations.

    Avoiding the double standard you correctly point out in #2 leads to a baby with the bathwater phenomenon. We want to argue correctly that CAM is a SCAM but we don't have much to offer some of the patients who find that sort of thing helpful. And we as a community have decided that "depression is a disease like diabetes" and that migraines should be called "migraine disease" even though that's not science based, and our treatments aren't that good. And so it's little wonder that patients who've continued to suffer after seeing a doctor or 3 are open to going to a craniosacral therapist and having their skull plates rearranged.

    A more honest approach would be to say that we have specific treatments for specific tissue diseases (a long list) but that we only have nonspecific treatments to offer for nonspecific conditions not explained by a specific mechanism (including much of chronic pain, chronic mild psychiatric symptoms, etc) -- this is the old disease vs illness distinction, which I think is still very useful, but which has sadly fallen out of favour. CAM should be banned from making specific treatment claims for curing diseases, though they might certainly help people feel better (I have no objection to someone saying their magnetic bracelet helps them manage their chronic rheumatoid arthritis pain, as long as we're not supposed to believe that joints are repairing themselves by magnetic fields, no matter how useful the illusion). For illnesses, including subjective overlays of objective conditions, CAM could compete equally with SBM -- as in fact it already does.

    (continued again...)

  6. (last part!)

    Your 3 points are right, but I'd qualify them as follows:
    1. placebos shouldn't claim specific effectiveness for diseases. Unfortunately, some of their nonspecific benefit (for illnesses) is likely from exactly that form of overpromising. I doubt many patients and caregivers will respond happily to instructions that "we all stop kidding ourselves". Yes I've read Kaptchuk on ethical placebos, but it's hard to imagine reworking much of modern medicine in that image.
    2. The double standard you identify is real, and truly hypocritical. I'm not sure it's avoidable though. There's a long philosophical tradition (dating back to Plato) of the societal value of a Noble Lie. It may be human nature to require a just-so story explaining ill health and treatment. If we cede the imaginative territory, the result will be more CAM, not more clear-headed acceptance of inevitable suffering.
    3. It is unscientific, yes. But medical care is fundamentally about caring for people. Science is a means for that (a hugely important means! our only hope for improvement! -- not being ironic with those exclamation marks) but it is not the whole story.

    Bottom line, I'm not so sure that "kidding ourselves" is entirely a bad thing, pitted against relief of suffering. There's a fault line between the cases where false placebo promises are cruel and deceptive and those where they are likely no worse than the best available treatment, and that's something that we might have to sort out as a profession, sooner or later.

    The way in which that might play out is with cost control. Part of the history of doctors amusing the patient is that our means of diagnosis and treatment have become increasingly technological and expensive. Even for chronic tennis elbow, not on anyone's list of the most complicated medical conditions, standard of care in the 21st century might include an MRI, injections of platelet rich plasma, 6 scheduled visits with a physiotherapist, and a consultation with a subspecialist orthopaedic surgeon -- especially if most or all of those options are covered by public or private insurance, or if there is a work related injury claim at stake. Total costs might well run into the thousands, even outside the US. All scientific and philosophical objections aside, if I'm right that many of the patients with prolonged disabling symptoms have a chronic pain disorder not well explained by tissue pathology, it's hard to imagine that such a comprehensive and expensive approach to managing the case is cost effective.

    The logic of medical care has been continual escalation, a one-way ratchet. It will always have been X months or years since the last scan or consultation, chronic pain patients presenting to clinic often describe their pain as worse recently, and pain is often associated with confusing behavioral overlays including gait difficulty and falls, so the logic will always exist to do one more scan and see one more specialist. It's hard to imagine that that's sustainable. Once every chronic migraineur (maybe 2% of the world) needs Botox (which is unarguably the best standard of care by the evidence right now), how are we going to pay for that?

    Medicine has built an empire extending into almost everyone's life, including miraculously effective diagnoses and treatments, but also many things we have little power to explain or cure. Some of that territory is built on a shaky foundation, and we will have trouble defending it long term.

    (sorry again for the length, maybe I need my own blog)

    1. Excellent comments, thanks.

      You could write a book, not just a blog.

      I agree with your points, but I see that we are coming at it from different directions. With migraine, you are probably seeing lots of non-specific treatments that are making patients happy, not costing too much and not really adding a significant risk of harm. Coming from surgery, I am seeing things like total hip replacements, sacro-iliac fusions and spine fusions performed on patients with normal X-rays and chronic pain syndromes who have had MRIs, scopes, physio, injections, and are on maximum doses of gabapentin and opioids. Surgery becomes the “next step” in the escalation of their treatment. Last resort, sure, but it should never be considered for these patients; the risk is too high.

      And if you think Botox is unaffordable, what about nearly half a million spine fusions per year in the USA, largely for “degenerative” conditions?

      Regarding CAM, I agree with you. My stance though, is to step away from the mainstream skeptics and science-based medicine community because their agenda is to bring CAM to account; mine isn’t. Not because I support CAM, but because of the double standard: if we are to apply these scientific principles to CAM, then we should also be applying them to medicine. I guess that is the niche that I am trying to fill - there are enough people putting the boot into CAM already. I think we are too busy throwing stones sometimes to look in the mirror.

  7. Different experiences, sure, but I agree with you completely that the risk (let along cost) is too high for surgery in the cases you mention. The problem is the incentives push doctors and patients in the wrong direction. As Anon pointed out upthread people will grasp at straws if we give them half a chance. Surgeons need to have the confidence (and the scientific evidence) not to offer the surgery in the first place. Trouble is, there will always be a new treatment or procedure showing impressive effects in a small open label case series, and by the time the RCT is done there will be true believers among patients and doctors alike. We've seen this in neurology over the last few years with the so-called "liberation" treatment for MS.

    I also completely agree with you in terms of the anti-CAM beat, which is already well covered by SBM and many others. I'm sympathetic to their work but find they tend to avoid some of the harder questions about the limitations of what science based medicine can actually offer people.

    Thanks for this exchange, it's great fun bouncing ideas off someone so clearheaded. I keep meaning to force myself to start writing about these sorts of things. I'll send you the link if I ever get around to setting up a blog.

    1. Too true: incentives for procedural doctors are high, and the ability/chutzpah to advise against intervention is lacking (inserting plug for recent blog post here:

      Thanks for your comments.

  8. I think you bring up a great point about CAM offering a better 'story' than mainstream biomedicine (physicians, physical therapists).

    You alluded to the pain science earlier, and I think that, plus the ideas of shifting to 'right-brain thinking' (global picture over small parts) we can start to offer our patients a better 'story.' Relevant metaphors with our newer, less-wrong scientific explanations can offer great benefits for the patients suffering with non-specific pain (i.e. a large portion of the persons utilizing health care resources).

    These patients deserve and require empowerment, empathy, knowledge, and no shame (this is probably one of the most important factors IMO) for their present state. Once we 'mainstreamers' can offer this, I think that most CAM will have no leg to stand on.

    1. Thanks Nick, you raise some interesting points about shame and knowledge. I am certainly a big believer in informing the patients as much as possible and having them contribute to the decision making as much as possible.

      I also agree that non-specific pain makes up a big part of health resource consumption, but doctors still want to reduce it to one spot on a scan and one targeted procedure, and are surprised when it doesn't work.

    2. I presume you are a doctor? I am a physical therapist. Many therapists are biomechanically oriented and operate in the dualistic world of Neuro vs. Ortho. They, as as you say, reduce this to one spot (not on scan, but with the hands [highly unreliable, but that is another topic]).

      Personal anecdote warning: my mother suffered from 'fibromyalgia' and chronic back pain for nearly 15 years. All throughout school, I went through all of the biomechanics/exercises/blah blah blah with her. None of it helped. Physicians just put her on ever increasing meds. I learn about pain science and teach her it, and it is almost as if the fibromyalgia 'disappeared' and she is becoming ever more active and happy. Within a ~month! Not only was the pain science a good story, but it was not rooted in pseudoscience.

      Now, as a physician I understand you have very little time you can spend with the patient. Talking them through their crisis and giving them some gentle manual contact to soothe their nervous system is probably low on the list below ruling out scary stuff. The incentive (at least in the U.S) is for you to not talk with them very long. How do you see this changing? How do you propose enlightening colleagues who are so rooted in biomedical approach?

    3. Thanks Nick,

      To change the system you would need to change the incentives. I get paid to operate, and earn less when I consult longer.

      Ideally we should be paid for results, and incentives that bias certain treatments should be removed. Easier said than done.

  9. Easier said than done, indeed. I try my best to remain optimistic, and educate/promote skepticism/scientific thought.

    But, the more silliness I hear, the more the cynic grows in me. Society (at least in the U.S.) seems to be on a crash course.

    It certainly is going to be an exciting time fighting the uphill battle against pseudoscience, bad incentives, and biomedicism (in the areas where it has done nothing but harm).

    I appreciate the discourse and wish you well in your endeavors.

  10. Elbow Tendinitis:
    may I suggest a read of the last 2.5 pages of Colin Campbell's book "The China Study"
    from half way down page 348.
    In 1988 my achilles tendons at age 43 had lumps for 4 years.
    Was marathon running sub 3 hour.
    I had episodes of plantar fasciitis, tennis elbow, shoulder tendinitis, hamstring tendinitis, iliotibial band inflammation at the knee (diagnosed by a physio when I thought it was lateral meniscal).
    Change of family circumstances caused me to be a single parent with one to three kids.
    Slowly on a largely plant based food program my tendinitis healed, minimal since.
    I expend considerable energy in running clubs, cycling clubs, walking clubs in promoting what I call working with nature's healing process rather than CAM or Integrative. It is hard to say much in my clinics which are medico legal and part time.
    Living with a waistline less than 90 cm for men and 80 cm for women (mine's sadly 95cm), being very active, meditating, is my line of thought.
    Listened to cardiologist Ross Walker on the Switzer program talking along similar lines in early May 2013.
    I recommend in it's general direction George Jelinek's book "Overcoming Multiple Sclerosis."
    I'm interested in the 10% of 4,300 centenarians in Australia who are in reasonable health.
    When patients are aligned with my way of thinking, I occasionally put it to them that
    tendinitis/arthritis can be a warning sign to improve nutrition, activity - or it's heart problems, diabetes, cancer ahead.
    Never came across much tendinitis working in Asia for a year, in Africa for 6 years.
    In Perth we have a group of up to 10 over age 60 who run 20 km at 1630 every Wed if anyone wants to join ..... we do it on mostly plant food.
    Just perplexed that in your elbow tendinitis writings there's not much on food, not much on lowering urate and Chol levels. Am I nuts ? John B age 67.

    1. Thanks John,
      Some interesting comments, but to answer your question specifically, I did not mention diet as a treatment for tennis elbow because there have been no experiments reported that determine the effect. THere may be anecdotal evidence, but if there are comparative trials out there, I would be happy to include them.

  11. Of course diet changes that the patient believe to be healthy can be a placebo too...

  12. Just came across your blog read it, I liked the information you have shared nice one...