Myths and misconception in Physiotherapy

Welcome to our list of myths and misconceptions in physio! This list is designed for students, but it might be useful for qualified physios, too. There are six sections of myths and misconceptions: assessment, treatment, clinical reasoning, condition-specific, exercise, pain science, and evidence based practice.

Why did I write this list? Well, as students, my friends and I spent a lot of time and energy working this stuff out. I want to save future students from having to do all that frustrating work, so they can concentrate on learning new things.

Reading the list, notice there are recurring underlying causes for these errors: overconfidence in what we can know or do, accepting the first explanation and never looking for an alternative, biomechanical and soft-tissue-based thinking, professional promotion, accepting society’s assumptions as our own, etc.

And don’t feel disheartened when you read these! All disciplines, for example medicinepsychology and education, have their own myths and misconceptions. “The greatest contribution to knowledge consists in removing what we think is wrong”. 

-Tom (@thomas_jesson)

Assessment myths and misconceptions
Myth Explanation
Postural, structural and biomechanical quirks show you the cause of pain Structure is not destiny! Pelvic tilt, lordosis, kyphosis, ‘over’ pronation, weak muscles, tight muscles, bulging discs, asymmetries, mild scoliosis: these things do not explain pain and worrying about can feed in to more pain.
Postural types are scientific They’re really pseudo-scientific, and posture correction has a long and ignoble history. The idea that we should sit and stand up straight is simply made up
Scans show you the cause of pain Treat the man, not the scan! Brinjikji et al found that people with no pain have all sorts of things going on in their spines. This doesn’t mean the scan is irrelevant, just that it is part of a complex picture.
Core instability causes back pain The one myth to rule them all. Peter O’Sullivan produced some of the research that started the core stability myth, and now totally refutes its value as an explanation of LBP. His youtube videos show his new approach.
Ben Smith’s systematic review is the most recent of many to indicate that core stability training is no better than any other exercise for LBP. This podcast is a nice intro to his work. 
Telling people their core needs stabilising is just plain mean.
Overpronation causes (insert pain here) “Overpronation” is everywhere, although there’s not really a consensus about what it is. Once spotted, it’s easy to follow a spurious line of reasoning to identify it as the cause of someone’s pain. But there are a lot of holes in this approach.
Palpation is a super-skill Gone are the days when physios believed we could detect meaningful information about the state of soft tissues or the ‘alignment’ of the body with our super-sensitive, magic hands. It seems our sense of touch is particularly vulnerable to over-interpretation.
Palpation is an illusion Is this in danger of becoming a new myth? Good hands are important for performing certain special tests, for manual therapy, and for communicating with your patient and making them feel safe. We are one of the only professions that gets to use touch to reduce people’s pain, help them move and build their strength and confidence.
Dermatomes are strict and predictable Dermatome maps don’t properly show the variation and overlap of our dermatomes.
Trigger points are (insert explanation here) No doubt something is there, although some experienced clinicians claim they have never felt one, and it’s a bit implausible that they are what many people think they are. Stay sceptical.
Treatment myths and misconceptions
Myth Explanation
Massage increases blood flow and has all sorts of effects on soft tissues Massage probably does increase superficial, cutaneous blood flow (making skin look red). But it’s unlikely it does anything to the blood flow of deeper tissues.
Massage is great but probably not because it is causing meaningful changes in soft tissues such as lengthening or aligning them, or washing out inflammation.
Manual therapy works by “pain gate” theory This theory was put forward by Melzack and Wall in 1965 and is very significant in the history of pain science because it was the first theory to account for top-down modulation of pain (ironically the bit that is most forgotten when people think of Pain Gate now). But it is such a small part of a complex picture that some time ago Melzack and Wall encouraged people to move on.
Joint mobilisations/manips put the joint back in place The mechanisms of manual therapy are mostly “neurophysiological”. Whether or not “neurophysiological” mechanisms are specific to manual therapy is up for debate.
Manual therapy is direction specific All pressure applied onto the skin is transferred perpendicularly through bone and soft tissue because the skin-fascia interface is frictionless. Here is an adorable video to explain this. This means that pulling, pushing and grabbing anything except skin in any particular direction is not going to happen.
Ultrasound causes cavitation and acoustic streaming The lack of effectiveness of ultrasound is well known, but less well known is that its proposed mechanism of effect, that it causes cavitation and acoustic streaming in tissues, may not even happen in living organisms.
Treatments work by causing soft tissue changes People often get better, but tissue stays the same. People who improve after core stability exercises, for example, don’t show any actual increased “core stability”. When tendon pain goes, the tendon has not “healed”. This is no surprise: loading is important but works in many ways, not just by causing structural changes. This is a good thing, it means exercise works in more ways than we think.
People with persistent pain should “pace themselves” Sometimes we tell people in pain to “pace yourself”. Often, this is heard as “do what you feel able to do”. In fact, pacing is about doing the same, manageable amount daily, regardless of how you feel, and increasing over time.
Clinical reasoning myths and misconceptions
Myth Explanation
If the patient got better, the treatment worked Why do ineffective treatments seem helpfulNatural history, regression to the mean, placebo, confirmation bias…
Pain is complicated Pain isn’t complicated, like a machine, where causation is predictable and knowable; it’s complex, like a city or an ecosystem, where causation is unpredictable and hard to trace. If one thing changes we don’t know what else will change and in what way. This means often we are fooling ourselves when we identify one or two original causes.
Special tests are special Special tests are not litmus tests: they don’t give yes or no answers. Not many have great sensitivity or specificity, and these values are merely estimates anyway. Also, the spPIN and snNOUT rule is a bit misleading because the rule-in and rule-out ability of tests depends on both their sensitivity and specificity.
Anything you can try to help the patient is good It is human nature to want to try every tool in the toolkit to help someone in pain. But often this approach forgets that as well as potential benefits, all treatments have costs. This may be in terms of time and money, which is an important consideration if you work in the NHS, but more importantly they may be in the form of harm to the patient. All treatments have the potential to rob patients of their self-efficacy and to encourage maladaptive beliefs and behaviour.
There are responders and non-responders to treatments The idea that certain people “respond” to certain treatments is appealing, and it is nice to think that there is a perfect patient for each tool in your toolkit. Unfortunately, data suggest that for our underwhelming pain treatments, including exercise, there is no hidden group of responders.
Placebo is a powerful force that we need to leverage The debates about placebo are endless but it is sufficient to say that we should be wary of anyone saying that placebo is 1) powerful or 2) something we need to “leverage” with complicated extra interventions. The fact is that placebo is 1) a limited, unreliable, unpredictable thing and 2) the best part of placebo can be obtained through patient therapist interaction without the need for needles or machines.
Condition-specific myths and misconceptions
Myth Explanation
OA is wear and tear OA is a metabolically active process, so telling a patient they have “wear and repair” rather than “wear and tear” is not only less fear inducing but more truthful.
Back pain increases with age Back pain does not become more prevalent after middle age, and may even decrease in the very elderly.
For patient who needs to hydrate, tea and juice are not good options Tea and juice are in fact more hydrating than water, so it’s okay for people with, for example, bronchiectasis to drink these things.
Shoulder impingement is a thing It’s not a thing. Neer’s model of shoulder impingement doesn’t really hold up and might be fear-inducing for patients. Is it better to say that weak and painful shoulders have “rotator cuff tendinopathy”? Any ‘impingement’, in this case, is secondary and not a diagnosis.
Running causes OA Evidence is conflicting and common sense suggests that well-managed loading is good for joints, so we can challenge patients’ beliefs that running causes “wear and tear”.
Tendinopathy has an inflammatory component It’s no longer appropriate to say “tendinitis”, or to manage tendinopathies like they are inflammatory.
Tendinopathy doesn’t have an inflammatory component Ok, there is someinflammation in tendinopathy!
There are adhesions in adhesive capsulitis There aren’t adhesions, so it might be better to call this condition “frozen shoulder contracture syndrome”. And since there are no adhesions to break down, it makes little sense to bring patients in for aggressive range-of-motion exercises.
Exercise myths and misconceptions
Myth Explanation
Stretching causes muscles to lengthen This is possible but takes a lot of force and time, and it’s unlikely that even a good long yoga session is enough to stimulate actual tissue changes – even in terms of tissue flexibility. Instead what is likely to be adapting is the nervous system, which is ‘reassured’ by stretching that the end range of motion is safe, and therefore ‘allows’ the muscle to lengthen a bit more next time.  **EDIT** Is this myth a myth? Update coming!
Stretching is good for tendinopathy Compression is a big factor in tendinopathy and many stretches compress tendons, if not against nearby bone then at least within their tendinous sheaths. So stretching may exacerbate tendinopathies.
Stretching reduces injury Stretching does not reduce sports injuries.
You can stretch ITB ITB is tough and firmly attached to the femur.
You shouldn’t squat knees over toes Unless there is a compelling reason, it’s fine for knees to go over toes, and for many squat patterns it’s essential. It’s going to happen when someone walks upstairs anyway!
VMO isolation exercises are useful for PFP All four quads are innervated by one nerve, so preferential activation for VMO is very unlikely and has never been shown before.
“Glutes not firing” as a cause of whatever problem It’s a bit mysterious where this idea comes from. Bits of your body don’t just fall asleep?
Gym ball exercises increase strength You can’t fire a cannon from a canoe!
Pain science myths and misconceptions
Myth Explanation
Pain is an issue in the tissues The now well-established disconnect between “hurt” and “harm” tells us that pain does not “come from” the tissues, and nor is it a good measure of the state of the tissues. Pain is an output of the brain, which scrutinizes many inputs, internal and external, including but not limited to nociception.
Pain is in the brain Some physios have accepted that pain isn’t in the tissues and decided it must be in the brain instead. But, it’s not there either. Whether or not saying “pain is in the brain” is a helpful thing to say to patients is debatable.
There are “pain signals” that travel along “pain fibers” “Eyes have light receptors, not vision. Ears have vibration receptors, not hearing. Tissue has danger receptors, not pain” — Adrian Louw. Pain is an output of the brain.
People in persistent pain have got central sensitization Central sensitization, the long-term potentiation of pain in the spinal cord and brain, is often assumed to be inevitable if someone has been in pain for a long time. But central sensitization isn’t just another word for persistent pain, it’s its own particular thing. So, someone with chronic low back pain may or may not have central sensitizationThis story illustrates one unexpected danger of assuming persistent pain is centrally driven.
Explaining pain is about teaching people to manage pain There are many educational interventions that help people to cope with pain, but pain neuroscience education is about reducing pain.
Pain science relegates the “bio” from the “biopsychosocial”. Pain science does not imply that bio-factors, including biomechanics, are irrelevant – it just puts them in a better context. Indeed, pain neuroscience education has not yet been shown in an RCT to reduce pain without some kind of exercise or manual therapy intervention to go with it.
Pain science is for persistent pain. Many patients can benefit from the message that hurt does not equal harm.
Evidence based practice myths and misconceptions
Myth Explanation
Level I evidence tells us what we should or shouldn’t do It would be nice to think that an RCT has the authority to tell us the truth about our practice. But research evidence is deeply messy and flawed. Most people are aware of this to a certain extent, but probably not the whole story. About half of published physiotherapy trials are false, and most of the quality indicators we learn about in textbooks – such as ITT analysis, sample size and blinding – don’t actually predict which ones. This is true for medical research, too. Not to mention the many ways bias can seep in to a trial.
A high PEdro score means a good paper PEDro is pretty good but it’s a blunt tool. For example, an RCT can get a high score but still be a meaningless A vs A+B study
There is a pyramidal hierarchy of evidence, with RCTs at the top EBM has moved on from the pyramid of evidence and developed more sophisticated models like GRADE. Good cohort and case-control studies can be as useful as a flawed RCT.
A systematic review is as good as it gets Garbage in, garbage out. If a SR uses bad primary research, the SR is bad. Same goes for meta-analyses.
Statistical significance is important for the clinic A result can be statistically significant but meaningless for your patient. Look at the raw data, the effect size, minimal clinically important difference, and so on. And raise an eyebrow whenever a trial’s results are simply reported in, say, a facebook post as “significant”.
The P-value is the probability that your results occurred by chance The P-value is the likelihood that chance could throw up your results even if there was no difference between treatments. The problem with this is that, in most things, it’s far more likely that there is indeed no difference than a real difference, meaning that you get a lot more false positives than you might expect. Add to this the fact that most studies are powered to find only 80% of true positives anyway, and false-positives hugely outweigh true ones. This is called the base rate fallacy. For example, even in a trial where the probability of there being a real effect is 50:50, a very high estimate, “P=<0.05” means that the false-positive rate is 26%.
A sample needs to be representative for a study to be relevant Commonly, you will hear results of trials dismissed because the sample does not represent a clinical population. While it is true that a representative sample helps to generalise findings, it can also undermine them. Representative samples have lots of confounding variables that reduce the internal validity of a trial. This means that external validity and internal validity become a see-saw. If a trial appears weak because its sample is not representative, it may be that its greater internal validity more than makes up for this.  
Correlation implies causation The number of people who drown by falling into a swimming pool correlates with Nicholas Cage movie releases
Evidence-based practice is like a stool The stool metaphor implies that all three legs – research, clinical experience and patient preference – have equal value when making clinical decisions. But they don’t.
Small studies are unlikely to have false positives. It is well known that small studies are worse at detecting true effects; that is, they make more type II errors. Often people take this to mean that if small studies do show an effect then that effect is more likely to be real. In fact, small studies are also more likely to show a false effect; that is, they make more type I errors, proportionately, than large studies, because there aren’t as many real effects to water them down.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s