How many shoulder classifications systems are enough?

Or rather, how many shoulder classification systems are too many?

Five sounds fair?

Ten seems a little bit too much?

and 18 would be… ridiculous right? RIGHT?!

Well folks, hold onto your shoulders.

In 2011 (15 years ago) a systematic review identified 18, yes EIGHTEEN, published classification systems in the existing literature, BEFORE they went on to develop their own. So 19 then. [1]

Over the years that number has continued to rise. Has this abundance of classification systems helped to improve the lives of people with shoulder instability? No.

Have the proposed rehabilitation regimes and treatment plans provided definitive answers on the best way to help all our patients? Not quite yet, more data needed.

Do I have all the answers? No.

Am I having a go? Definitely not.

Shoulder instability is a small world. My experience so far in this space has been positive. This is because those who I’ve engaged with have been welcoming, receptive to new ideas, giving of their time and willing to engage in the conversations. (Thanks to all if I’ve not said so already!) They are often very experienced and knowledgeable clinicians all of whom are trying to do what’s best for people with shoulder instability.

It’s important to remember that a lot of the concepts critiqued here came about as a result of trying to make sense of the chaos. They were the original ideas, innovations, fledgling hypotheses and the foundations from which other concepts could develop and compare themselves. Standing on the shoulders of giants and all that.

History tells us that our initial ideas of how the world works aren’t always correct the first time around, (think flat earth theory, milk blood transfusions, and cocaine for hay fever [2]). We need to continually revisit our assumptions and understanding. It’s difficult to do this when there’s no new or reliable information.

It seems like every time I get online, someone, somewhere has published a new shoulder instability classification system, rehabilitation plan or course.

I’m not against that. This is a complicated area. Anything that brings about more clarity and certainty can only be a good thing. But we have to be honest about the state of our clinical practice, the evidence that drives it, and whether these growing contributions are actually helping or even making things worse. Is the information really ‘new’? Is it fit for purpose?

To help answer this, in this article I want to put forward three main points:

* Existing classification systems are conceptually useful but practically useless.

* Research in this area is thin, not well joined up and does not follow the life of the patient.

* We seem to be repeating the same mistakes and it’s time to try something different.

Point 1 is the longest. If you make it through that, you’re on the home straight.

Why is this important?

There is a phenomenon in artificial intelligence (AI) known as iterative degradation or quality decay. You may have seen funny examples of it on the internet where ChatGPT is asked to create exact replicas of a picture over and over again. The final result ends up worse than the original (Figure 1). The cause of the problem comes from the fact that every time a ‘new updated version’ is fed back into the model, small errors and noisy data get larger.

Figure 1. ChatGPT is asked to make exact replicas of a picture of Dwayne ‘The Rock’ Johnson 101 times. Source: https://www.reddit.com/r/ChatGPT/comments/1kbj71z/i_tried_the_create_the_exact_replica_of_this/

Without new or accurate information to correct itself against, the system just gradually becomes a parody of its former self. Can you see where I am going with this?

At the patient level, this potentially results in them not getting the best care. Misdiagnosis and wrongfully attributing things as the cause of their problems may result in delays or withholding of needed investigations or treatments. It may also result in unnecessary or ineffective treatments being used, all of which make the overall healthcare burden on people larger and longer.

At the therapist level it makes entering the world of shoulder instability seem overly complex. As an exercise, I always ask myself, if I were a student trying to learn about this afresh, and all I had available to me was the published evidence, would I arrive at the same conclusions as someone else?

Is the data or evidence clear enough?

If I had to pick, which classification system or treatment plan would I use and why?

How much of current practice is a product of tacit knowledge, departmental or institutional norms? More recently social media, course and masterclass conference presentations are playing a larger role and so we need to make sure we can help people discern the truth.

So here’s my attempt at doing that.

1. Existing classification systems are conceptually useful but practically useless.

The purpose of classification is to identify distinct groups to which people or things can be assigned on the basis of some predetermined measures or characteristics. For example in Figure 2 you could classify Lego blocks by their colour or their shape.

Figure 2. Lego blocks arranged by colour

Photo by Mourizal Zativa on Unsplash

These features (shape and colour) are practical, cheap and easy to observe (measure). You could argue that shape would be a more robust method for classifying Lego blocks. This is because misclassification may occur by people who are colour blind as in Figure 3. The main point here however, is that any object could be classified in multiple ways depending on the measure or feature, or combination of measures and features used.

Figure 3. Examples of changes in colour depending on type of colour blindness

Image taken from https://midtownvision.com/blog-posts/types-color-blindness

Classification systems are therefore only as good as the measures on which they are based. If the measures or features aren’t reliable, are subject to interpretation or can’t be quantified, it may result in misclassification (like in the picture above).

We also need to consider, did we make the classification systems and then assign measures we think should best go with them (top down)? Or, did we let the data and its features objectively (and without bias) tell us if there are actually distinct or overlapping groups (bottom up)? My feeling is that in healthcare we too often do the former.

Another thing to consider is what purpose does the classification system we are using serve? Is it a conceptual and theoretical model to help us make sense of an incomplete understanding of physiological or biomechanical processes? Or, is it used as a definitive treatment algorithm based on objectively quantified measures or tests?

Are we guilty of confusing the two?

Do we erroneously give equal weight to both when we shouldn’t?

So what is actually needed for a classification system to be any good?

* The end treatment, investigation or management plan has to be different.

* If all people irrespective of category get the same investigations or treatments, is the classification system serving a purpose?

* It has to put people in distinct groups i.e. it has to be able to discriminate.

* What if a person can simultaneously exist across all possible categories, each of which has a different investigation or treatment endpoint? Does the classification system have enough discriminatory ability to be considered suitable for practice?

* It needs reliable and accurate measurements.

* People need to agree on the way these measurements are used to combine or group people.

So do existing shoulder instability clinical classification systems work? Not really, especially when compared against imaging or surgery which is usually considered the ‘gold standard’.

Here are some quick stats:

* Moroder et al, 2020 [3] reported that multidirectional instability was over diagnosed. 10 to 20% of patients had bony changes in their shoulder despite the cause of their shoulder problem being a ‘muscle co-ordination issue’.

* Jaggi et al, 2023 [4] found that 10% of participants were later found to be unsuitable for the study after an arthroscopy showed no capsulolabral damage or a bony injury.

* Clarke et al, 2024 [5] - summarised nicely by Adrian Davies [6] also shows an approximate error rate of 20% as well when it comes to confirming the direction of instability (in rugby players).

Note: I have also purposefully mixed ‘traumatic and atraumatic’ instability and go interchangeably at times between them. The misclassification problem is a reason for this. When I refer to shoulder instability I am talking about the less obvious cases, although the concept could be extended to any subgroup.

Now granted, there are some barn door obvious cases which arguably don’t need a classification system e.g. some traumatic dislocations accompanied by imaging. But as a clinician I’m usually more interested in the ones we don’t get right or are unclear, rather than the ones that are obvious. It appears we are getting it wrong for about 1 in every 10 (if you want to be optimistic) or 1 in 5 patients. Is that good enough?

Why is this happening? What is it that existing classification systems aren’t telling us? To help illustrate some of the points I’m going to use the Stanmore Triangle.

The Stanmore Triangle (reproduced below) frames shoulder instability as a dynamic or shifting condition, existing between 3 poles and 5 states (structural, non-structural, traumatic, atraumatic and muscle patterning - can you see why I said this may be overly complex?). The idea is that people can move along these poles and between states with varying levels of each state (basically along the lines of the triangle and across to other corners or poles).

Reproduced from Lewis, A., Kitamura, T., & Bayley, J. I. L. (2004). (ii) The classification of shoulder instability: new light through old windows!. Current Orthopaedics, 18(2), 97-108. https://doi.org/10.1016/j.cuor.2004.04.002

Seems like a handy framework right? But what happens when we get into the nuts and the bolts of it? Can it work practically? Does it do all of those things that are needed for a classification system to be any good? Does it put people in mutually exclusive categories? If the management options are different, but you can’t clearly identify who needs what accurately then what purpose does it serve?

It took me a while to recognise where else I had seen this model before. It’s the same one used in theology to depict the Holy Trinity and it’s been keeping theologians busy for thousands of years as they depict (without committing heresy) an entity that is simultaneously all things at once and yet three distinctly different things at the same time. Somehow we’ve decided it’s a good model for shoulder instability. I’ve also seen people use triangle models to describe states of selective emotions e.g. excitement, nervousness and cynicism for an upcoming event. A triangle model in this case also makes sense. Having a combination of different emotions and transitioning between them regularly seems a conceptually helpful way to illustrate how someone feels.

So it seems triangle models are helpful for concepts but not necessarily clinical decision making. At the end of the day we are therapists not theologians, practitioners not philosophers.

There’s a few more things we need to consider. Irrespective of the shape (although I have seen someone propose the cube of shoulder instability to try to resolve this), the measures on which the classification is based are arguably the most important. For example, what does muscle patterning actually mean? Are we talking about a co-ordination issue? How would we detect this? What if the muscle patterning is a consequence of changes to the underlying structure rather than the cause. As things stand there’s not enough high quality evidence (based on state of the art 3D and surface electromyography measures) to differentiate between movement patterns that could be considered:

* normal variation within a population

* different but an adaptation for maintaining stability based on changes to the joint from instability and

* different and is resulting / causing instability.

If we can’t yet make this distinction with equipment that allows us to go back and look at the different joints and muscles at a pace we can comprehend and revisit, how then can we do this in real time, with our eyeballs during clinic?

There are of course other classification systems available. The Frequency Etiology, Direction and Severity (FEDS) system is a helpful way of standardising the description of someone’s instability. Again, it seems pretty straightforward on the surface, however, when you look a bit more closely there are a lot of ways (30 different combinations) that people can be classified. It also doesn’t really tell us about the primary mechanism. Etiology is discussed in relation to being traumatic or atraumatic but with no reference to what role bony, muscular or soft tissue structures may play.

A lot of the models also do not explicitly mention what role psychosocial factors play. As a result it seems people will borrow categories or attributes from other models creating a sort of Frankenstein’s monster classification system that may have negative downstream effects. For example, as a precursor to the Stanmore Triangle, in 1979 Rockwood identified that the presence of psychiatric problems was relevant for people who could voluntarily sublux their shoulder in type 3 instabilities. Whilst not explicitly stated in the Stanmore Triangle this historical association has continued to permeate into practice. The end result? Some of our research that used hypothetical patients to see how physiotherapists make decisions for shoulder instability patients identified that 1) female patients were more likely to have negative psychosocial factors attributed to them (despite having none stated in the cases) and 2) male participants were more likely to be offered investigations sooner [7].

Stop here and consider the practicalities and real world implications of this.

How can we practically and objectively identify people with genuine psychosocial and psychiatric factors that are contributing to their shoulder instability? I mean what does contributing in this sense even mean? What is the mechanism by which these will impact the stability of someone’s joint beyond the pre-existing constraints of their underlying bony morphology, soft tissue structures and muscle architecture? The risk here is guilt by association. The mere presence of non-traumatic instability makes you look for psychosocial factors more intentionally than you would in other forms of shoulder instability.

If you were constantly worried about your shoulder coming out, or worse, it did keep coming out, your mood or mental health might be affected too. It’s also likely to be sore and limit what you can do which doesn’t help either. Next thing you know some therapist has decided the thing that’s most likely contributing to your shoulder instability is some form of life stress based on your gender although they don’t actually tell you this. We need to be careful about how and who we attach labels to, especially in the absence of any data we can point to and say “look here it is!” Classification systems that inform decision making are meant to provide objectivity in decision making by at least providing a check on our personal biases. They definitely aren’t meant to reinforce them.

So what’s the summary of all this? It seems existing models are incomplete. They can’t seem to identify and describe in unambiguous terms all of the important information required to accurately describe and classify all forms of shoulder instability. This may not be such a big problem if we used shoulder instability classification systems conceptually rather than practically.

But why are we in this cycle of inventing more and more classification systems? Well I’m glad you asked, the reason for this I believe is…

2. Research in this area is thin, not well joined up and does not follow the life of the patient.

Photo by engin akyurt on Unsplash

So what does the research tell us from a rehab perspective? Based on some (of the better) studies we know that:

* In adults with atraumatic surgery there is not much difference in outcomes between diagnostic arthroscopy (placebo) or actual arthroscopic capsular shift surgery [4].

* Existing uncertainties and a lack of robust evidence means that clinical recommendations were centred around expert opinion in arthroscopic shoulder stabilisation surgery for traumatic shoulder instability [8].

* In adults (general public, typically non-athletic sample) with a first time traumatic dislocation, additional/multiple sessions of physiotherapy were not superior to a single session of advice, supporting materials and the option to self refer to physio - The ARTISAN trial [9].

* High-load strengthening exercises are more effective than low-load strengthening in mainly adult females with hypermobile shoulders although results were variable [10].

* In adults with multidirectional instability the Watson multidirectional instability programme seems to perform better then the Rockwood Instability Programme [11].

* There is very little, if any, information available for young people with shoulder instability.

* Even in the basic science/mechanism space, I was surprised when writing up our study [12] that all of the previous research on muscle activity and movement differences were predominantly adults with the odd young person. Our dataset is probably one of the youngest recorded.

There are other studies that provide helpful information such as the Derby Instability Programme [13] and other similar single group, longitudinal studies. But remember when we want to know if something actually works or is better than something we are already doing, it’s usually well-designed and appropriately powered RCTs that are needed. Although there are other ways of evaluating effectiveness that I’ll discuss later.

Consider then how does the existing evidence inform our decision on what to do with a young person who has multi-directional/atraumatic shoulder instability or an overlay of traumatic and atraumatic? Where are the RCTs for them to inform our decision making?

The truth is getting funding in this area is difficult. Compared to other health conditions e.g. cancers, hypertension or heart attacks, shoulder instability as a proportion affects a small percentage of the population. It therefore doesn’t always tick the ‘value for money’ box funders are after. It also means getting enough people through the doors to answer the study question is more challenging.

It’s also partly a hidden problem. Estimating the true incidences for some of the more complex shoulder instability subgroups is challenging due to a lack of robust evidence (a recurring theme in this field). No one really knows how big the problem is and the complications are often delayed. Funders also usually want to fund studies that answer a question in a reasonable amount of time.

What’s not quantifiable in the data but very important to consider is that people with more mixed pictures/complex/multidirectional instability:

* Usually present as children or adolescents. They may have several episodes of instability and do not present until it starts impacting on them and their parents’ or carers’ daily lives in a significant way.

* Likely that the true size of the problem e.g. number of instability episodes and duration of problem is under reported.

* May be reliant on their parents or carers for communicating this information given their age.

* Their voice in the research agenda is therefore not prominent. They also likely have other life priorities or issues to sort out (remember how complex being a teenager was AND then your arm keeps popping out).

* Can spend a long time in the healthcare system going between services, despite being a relatively small proportion of service users.

* They can be complex cases for clinicians.

* They may see multiple clinicians and be referred between multiple services (high healthcare burden for them and us) which,

* may not result in joint up care. Throughout their life how many healthcare providers might someone with recurrent instability see (involves telling your story to lots of different people again and again)?

* Overall a worse experience of healthcare (having to wait for formal diagnosis and feeling like you are being passed between services. Consider also the potential for being labelled as having ‘psychosocial’ problems).

Unlike surgical interventions there’s not a culture / governance requirement to record and report outcomes. The end result is we can’t actually evaluate what we are doing to put a number to the problem.

There’s also a funding gap which I find a bit mind-boggling. We know that shoulder arthritis is a big problem. We know that in traumatic shoulder instability you are between 10 to 19 times more likely to get shoulder arthritis later on.

What we don’t know is how likely arthritis is to develop in someone with atraumatic shoulder instability who is experiencing multiple episodes of instability again and again.

There are charities/funders set up to address arthritis.

There are charities/funders set up to address health issues in children.

Both are at opposite ends of the spectrum, meaning those with the more complex and under-researched shoulder instability presentations fall between this gap.

If arthritis is such a big problem why aren’t we focusing on those most at risk to try and prevent it (if possible)? If funders are serious about addressing underserved groups or ensuring fairness though then surely this is one of the groups who should be considered?

So in essence the reason there’s limited new and rigorous evidence is that it’s pretty hard to get the big definitive studies off the ground for the reasons stated. Often if there isn’t a clear pathway to a big clinical trial that answers a specific clinical question, it can make it a bit more challenging to get the initial funding that’s required to develop these projects. Research studies therefore tend to focus on small development or pilot work which doesn’t always progress on. The end result is lots of research, not all of which can be used to inform our decision making.

Doing good research in this area is not impossible, just really really difficult. There’s also a circular logic problem. A lack of evidence makes getting funding to do more research difficult… which is needed to produce more evidence. The problem here seems to be…

3. We seem to be repeating the same mistakes and it’s time to try something different.

Photo by Collab Media on Unsplash

Albert Einstein is famous for saying “insanity is continuing to make shoulder instability classification systems and treatment plans in the absence of new data or RCTs” , or something like that.

Given that we’ve gone beyond 19 instability classification systems and the rehab evidence base is likely less than 10 papers it’s probably time we did something different.

The way I see it there are two possible options moving forward:

* This blog inspires a drastic change in how shoulder instability work is funded leading to more definitive trials (I wish), or

* We need to try something different, innovating and working within the systems we’ve got.

I think a fundamental step in this is improving how we measure people with shoulder instability, for both diagnosis and over their lifespan. Whilst RCTs are an excellent methodology for demonstrating effectiveness, they aren’t always as well suited (practically) to less common conditions or subgroups.

That doesn’t stop us from trying to answer some of our ongoing clinical questions using alternative methods, as long as we understand the implications of these. Remember in some cases we won’t be starting from a very high bar e.g. just expert opinion. There are some innovative studies going on in this space which provide hope:

* Moroder et al, 2020 [3] used fluoroscopy to classify and confirm different types of shoulder instability. I agree we can’t use that on everyone but it’s a helpful starting point from which other technologies or solutions may develop.

* In other rare diseases or complex movement disorders, things like 3D movement analysis and electromyography are used to help inform decision making. They also tend to have really good datasets that follow the patient throughout their life. Gillette Children’s Hospital is an excellent example of this.

You might say the technology is not there and it’s too complicated for practice. The truth is the technology has been there for quite a long time and is already used for making decisions about rehab and surgery. These arguments were around in the early days of human movement analysis itself and likely CT (just run that past me again… you want me to buy a £100,000+ machine that emits ionising radiation whilst spinning at 120 to 240 revolutions per minute and you have to lie in the middle of it?). However, as we were able to demonstrate how improved diagnosis improved outcomes for people, these new ways of diagnosing people were introduced and made standard in clinical pathways.

If we can improve our diagnosis alongside improved recording of outcomes that follow people throughout their life, we can then start to make positive steps to seeing not only how big the problem is, but what seems to change outcomes. This is always the starting point for further, well-designed definitive trials. In a world of digital health, secure data environments and AI, this should be easier than ever. However the risk is these just become really good systems that store and regurgitate really rubbish information. We need to try to avoid iterative degradation.

So what needs to happen next? My hope is that if nothing else, we stop and take stock of where we are at. The answer to too much unusable research is not more unusable research. The answer to too much complexity is not more complexity. We need to go back to basics. If we are to continue to keep rising while standing on the shoulders of giants, it’s important the foundations are right.

References

[1] Kuhn, J. E., Helmer, T. T., & Dunn, W. R. (2011). Development and reliability testing of the frequency, etiology, direction, and severity (FEDS) system for classifying glenohumeral instability. Journal of shoulder and elbow surgery, 20(4), 548-556.

[2] INGALS EF. COCAINE IN HAY FEVER. JAMA. 1886;VI(8):206. doi:10.1001/jama.1886.04250020066006

[3] Moroder, P., Danzinger, V., Maziak, N., Plachel, F., Pauly, S., Scheibel, M., & Minkus, M. (2020). Characteristics of functional shoulder instability. Journal of shoulder and elbow surgery, 29(1), 68-78.

[4] Jaggi, A., Herbert, R. D., Alexander, S., Majed, A., Butt, D., Higgs, D., ... & Ginn, K. A. (2023). Arthroscopic capsular shift surgery in patients with atraumatic shoulder joint instability: a randomised, placebo-controlled trial. British Journal of Sports Medicine, 57(23), 1484-1489.

[5] Clarke CJ, Torrance E, Gibson J, Brownson P, Funk L. Diagnosing the direction of shoulder instability in rugby players. Shoulder Elbow. 2024 Feb;16(1):33-37. doi: 10.1177/17585732221092025. Epub 2022 Mar 31. PMID: 38435041; PMCID: PMC10902408.

[6] Adrian Davies - Glenohumeral instability under the spotlight: Clinical findings vs. Intra-operative diagnosis, Adrians Shoulder Blog ttps://open.substack.com/pub/atdshoulder/p/glenohumeral-instability-under-the?r=37fnk4&utm_campaign=post&utm_medium=web

[6] Philp, F., Faux-Nightingale, A., Woolley, S., de Quincey, E., & Pandyan, A. (2022). Evaluating the clinical decision making of physiotherapists in the assessment and management of paediatric shoulder instability. Physiotherapy, 115, 46-57.

[7 Wong, C., Jaggi, A., Willmore, E., Maher, N., Bateman, M., O’Sullivan, J., ... & Chester, R. (2025). Critical evidence synthesis on rehabilitation following arthroscopic shoulder stabilisation surgery for traumatic anterior instability: consensus recommendations for clinical practice and research–commissioned by the British Elbow & Shoulder Society. British Journal of Sports Medicine.]

[8] Kearney, R. S., Ellard, D. R., Parsons, H., Haque, A., Mason, J., Nwankwo, H., ... & Underwood, M. (2024). Acute rehabilitation following traumatic anterior shoulder dislocation (ARTISAN): pragmatic, multicentre, randomised controlled trial. bmj, 384.

[9] Liaghat, B., Skou, S. T., Søndergaard, J., Boyle, E., Søgaard, K., & Juul-Kristensen, B. (2022). Short-term effectiveness of high-load compared with low-load strengthening exercise on self-reported function in patients with hypermobile shoulders: a randomised controlled trial. British Journal of Sports Medicine, 56(22), 1269-1276.

[10] Warby, S. A., Ford, J. J., Hahne, A. J., Watson, L., Balster, S., Lenssen, R., & Pizzari, T. (2018). Comparison of 2 exercise rehabilitation programs for multidirectional instability of the glenohumeral joint: a randomized controlled trial. The American journal of sports medicine, 46(1), 87-97.

[11] Seyres M, Postans N, Freeman R, Pandyan A, Chadwick EK, Philp F. Children and adolescents with all forms of shoulder instability demonstrate differences in their movement and muscle activity patterns when compared to age- and sex-matched controls. J Shoulder Elbow Surg. 2024 Sep;33(9):e478-e491. doi: 10.1016/j.jse.2024.01.043. Epub 2024 Mar 10. PMID: 38467183.

[13] Bateman, M., Osborne, S. E., & Smith, B. E. (2019). Physiotherapy treatment for atraumatic recurrent shoulder instability: updated results of the Derby Shoulder Instability Rehabilitation Programme. Journal of arthroscopy and joint surgery, 6(1), 35-41.

This summer, I was watching my two children running around the garden with their cousins. They don’t see each other very often so there was a lot of energy, a lot of noise, a lot of chaos, and a lot of joy. And as I watched them, it struck me that I could describe exactly the same scene in a few very different ways.

I could say they were a group of children.Or a gaggle of mischief makers.Or a gang of youths.

Nothing about the behaviour changes. But the words I choose tell you a lot about how I feel about what’s happening in my garden. And it’s exactly the same when we talk to patients about pain.

The (Chronic Low) Back Story

Chronic low back pain is associated with disability, work absence, unemployment, and high healthcare use. Quality of life is often dramatically reduced. In one study, people with chronic low back pain reported quality-of-life scores comparable to patients receiving palliative cancer care [1, 2]. Despite chronic low back pain being highly prevalent, it is still widely misunderstood. Large proportions of the general public believe back pain is caused by injury or damage, that bed rest is the best treatment, and that once you’ve ‘hurt your back’ it will never be the same again. What’s more concerning is that these beliefs are also common among healthcare professionals [3].

And that matters because clinicians play a major role in shaping how patients understand their pain. We often talk about challenging patient beliefs so they can engage better with treatment. But inaccurate beliefs don’t just affect behaviour. Inaccurate pain beliefs can increase threat, and increased threat increases pain.

Research shows that when clinicians deliver strongly biomedical messages, patients develop hypervigilance, more fear, and more disability. These patients often feel guilty if they can’t follow advice perfectly. And even when they do everything ‘right’, pain flares can feel baffling and demoralising. In contrast, when clinicians provide accurate reassurance and encourage movement and confidence, patients report higher self-efficacy, greater activity levels, and better engagement with life; not just exercise, but meaningful activities and social connection [4].

Pitfalls in Pain Science

Most clinicians can confidently define acute and chronic primary/secondary pain, or tell you the difference between pain aetiologies like nociceptive, neuropathic or nociplastic.

But even with this clinical knowledge there are some common pitfalls that clinicians can fall into when forming their own beliefs about pain.

Pitfall #1: Acute pain ‘goes chronic’ at 3 months

Acute pain is defined as pain that’s been present for less than three months, and is more likely to be associated with nociception. Chronic pain is defined as pain that lasts longer than three months. Pain doesn’t magically change character at two months and twenty-nine days. What that timeframe reflects is two things: the time most tissues take to heal, and the point at which a number of neurological mechanisms that can maintain pain tend to become established. That three-month mark isn’t arbitrary, but it also isn’t a hard line.

Not only that, but some chronic or persistent pain syndromes occur in the absence of an initial acute injury, and some are a result of ongoing nociception from a chronic disease.

Pitfall #2: That some pains are biological and some are psychological

When we talk about pain mechanisms, we usually describe three broad categories.

ACL injury and patellofemoral pain (PFP) are rarely invited into the same conversation. One is acute, traumatic and unmistakable. The other is persistent, frustrating and often labelled ‘non-specific’. Yet when we step back from tissues and diagnoses and instead consider risk, these two conditions begin to overlap in important ways.

Both are common. Both disrupt participation in sport and physical activity. And both emerge from a complex interaction between load, movement strategies, preparation, and behaviour. The difference is not how much we understand about risk — but how often that understanding meaningfully shapes what happens in practice.

Different injuries, similar environments

ACL injuries remain most common in sports involving cutting, pivoting and jumping, with match play carrying a significantly higher risk than training [1]. PFP, meanwhile, continues to be one of the most frequent knee presentations across adolescents and adults and is a leading cause of reduced physical activity and sport participation [2].

Despite obvious differences in presentation, both conditions tend to arise in similar environments: high training loads, repeated exposure to demanding tasks, limited recovery, and pressure to perform. From this perspective, ACL injury and PFP look less like unrelated problems and more like different outcomes within the same sporting and physical activity landscape.

Risk factors: real, but rarely simple

It is important to be clear about what we do not know. There is no single causal pathway for ACL injury. Rupture does not result from one faulty movement, one weak muscle, or one missed warm-up.

I think I know the MSKMag audience well enough to know that we’re a bunch that does love some evidence. We debate the nuances of treatment techniques, and we even (sometimes) read past the abstract of the RCTs we’re citing. We do this for a simple reason: we want to provide the best possible care based on what is actually happening on the ground rather than ‘gut feeling’ or outdated traditions.

Yet, when we step out of the treatment room and into the ‘business’, that rigour often vanishes. We operate our clinics by feel, doing what we think is best without truly knowing. It’s not entirely our fault either; there actually isn’t that much data to guide us. What is it that makes a successful clinic owner successful? What are the most profitable clinics doing differently? No one really knows because nobody has ever asked. So we’re instead left with gurus to guide us on the ‘business’ – the exact thing we abhor.

HMDG launched The Private Practice Barometer late last year with a simple aim: collect anonymous data from MSK clinic owners to finally answer some of those questions regarding the business side of the industry and the people that are running it. And despite the protestations from some who wrongly believed the intention for the survey was some secret other reason that they’d entirely made up in their heads - over 700 volunteered some data! And answers have been found! Huzzah!

Now I can’t leak all of the results yet - that will come soon, I promise. But what I can do is hand-pick some of the points that I think are of particular interest.

Clinic Ownership - Is it worth it?

I’ll start with the question that I was most intrigued to see the answer to. What is the daily reality for most clinic owners? Are they happy? Are they stressed? Are they too busy counting their cash to care? There is an aspirational narrative often sold to clinicians about opening your own clinic: that it brings complete clinical freedom and WEALTH! Surely? Well, it turns out – maybe not quite.

One uncomfortable truth staring out of the data is that the median MSK clinic owner earns around £48k annually - far below the compensation you would expect of any CEO. The reason I find it a particularly interesting data point is that the amount is also about the market rate of a senior clinician – in fact it’s less than a Band 7 salary in the NHS. And yet the responsibilities of running a clinic, along with the risk involved, comes with far more stress than simply managing a case-load provided for you.

Fibromyalgia is a symptom-based disorder characterised by persistent widespread pain and accompanied by fatigue, poor and unrefreshing sleep, low mood or depression, and cognitive symptoms. The diagnosis is made using the 2016 American College of Rheumatology criteria [1] and, importantly, is not a diagnosis of exclusion, which has significant clinical implications for assessment and management. The diagnostic criteria assess both the extent and spread of pain, as well as the presence and severity of associated symptoms including fatigue, waking unrefreshed, and cognitive symptoms. In addition, symptoms must have been present for three months or longer. Recognised risk factors include female sex, older age, increased body mass index, the presence of medical co-morbidities, a history of poor sleep, and exposure to stressful life events or depression [2]. Fibromyalgia is estimated to affect 2–3% of the general population [3]. The condition has wide-ranging effects on people’s daily lives and often results in reduced ability to engage in occupational and leisure activities, social isolation, and disrupted relationships [4]. More generally, widespread pain is associated with high pain intensity, increased healthcare utilisation [5], and a history of poor outcomes following multiple interventions [6]

At time of writing, I am on Day 1335 of learning French on Duolingo. I am very, very good at Duolingo. Unfortunately that doesn’t mean I’m very good at speaking French; in fact, put me in front of a French speaker and I’ll be asking them to slow down and perhaps even draw me a picture.

If you’ve not used the app yourself, let me introduce you to Duo, an exceptionally needy, green owl character who will ruin your life with constant reminders to refresh your language skills daily to maintain your streak and not let your app friends down horribly.

Any amount of your chosen language per day will keep Duo happy (though you will hurtle down the league tables if you don’t accumulate enough points that week).

The modules do incorporate varied topics and test you on reading, speaking, listening and writing language skills…but unless you pay even more for premium premium plus Duolingo, you have to figure out grammatical errors yourself because it won’t tell you exactly why your incorrect answer was wrong.

Sometimes I wonder if I’m learning French or just playing daily French-accented games as quickly as possible in order to win points, receive a pat on the back from a cartoon bird of prey and move on with my day, streak intact.

This month, Duolingo introduced ‘pranks’, the absolute cherry on top of features that are non-conducive to learning. By ‘pranking’ another user, you can have your avatar’s head pop up on their screen for a few moments, blocking their access to the lesson and slowing their progress (seemingly so that you can nip up the league table at their expense in the meantime). I was about to write that there’s a reason no one has thought to barge into an in-service training session, flap around for a bit, then wonder why no one picked anything up but then I remembered that some of us have actually worked for NHS departments.

I digress.

The reason Duolingo comes to mind this month is because it mirrors something we can do remarkably well in MSK practice: staying active (maintaining our streak) without always building towards progress.

This month’s Mag articles show us just how much work is being done to improve practice.

Derek Griffin’s Unfogging Fibromyalgia tackles a condition that remains widely misunderstood; getting to the bottom of the science behind it. Similarly, Julia Gover’s The Back Story asks us to look not at what we know about chronic low back pain, but at how we explain it, and whether our narratives genuinely empower patients. Elsewhere, complexity itself comes under scrutiny. Fraser Philp’s Shoulder Instability: It’s All Getting a Bit Silly highlights how an ever-expanding array of classification systems prove that more structure doesn’t always mean more clarity. Tom Jacobs’ ACL and PFP: Related Risk Factors challenges us to think beyond neat diagnostic silos, showing how interconnected problems often share roots that are easy to miss. And then there’s Mark Reid’s The Private Practice Barometer, which shows us the data as to how we are actually working and the impact that is having on private clinicians - because learning doesn’t just happen in journals.

Taken together, these articles represent what we have been doing well: questioning assumptions, revisiting explanations and trying to make sense of increasing complexity. But they also quietly expose where we could do better: by prioritising understanding over accumulation, coherence over completeness, and communication over mere correctness.

Because, as with my French, it’s one thing to keep the streak alive. It’s another to be truly fluent.

Chiropractic manipulation has been around long enough to be both revered and misunderstood in equal measure. For some, it’s the magic button that fixes everything from a sore back to a bad mood. For others, it’s a mysterious cracking ritual that must be avoided at all costs. In truth, spinal and extremity manipulation is a highly skilled, evidence-informed tool, but like any powerful tool, it requires wisdom, restraint, and a healthy dose of common sense.

Let’s start with a few common misconceptions. No, we don’t ‘put bones back in’. Your vertebra doesn’t go missing in action and need clicking back into place. And no, your spine is not now ‘aligned with the universe’. It’s a joint, not a solar system. And despite the internet’s enthusiasm, not every click, crack, or pop you see on social media is therapeutic; sometimes it’s just noise, clickbait, or a little harmless ASMR for the viewer and listener; best enjoyed with some doom-scrolling.

One of the reasons manipulation attracts such polarised opinions is that it produces an immediate, tangible sensation. Patients feel something happen. Clinicians hear something happen. In a healthcare world increasingly driven by slow-burn outcomes and delayed gratification, that immediacy can be both its greatest strength and its biggest trap. Humans are wired to associate instant feedback with effectiveness, even when long-term outcomes may tell a more nuanced story. A click can feel convincing; a quiet improvement over six weeks of rehab, less so. This doesn’t make manipulation ineffective, far from it, but it does mean we must be careful not to confuse impact with importance.

There’s also a performance element that’s hard to ignore. Manipulation looks confident. It looks decisive. It looks like you know what you’re doing, which is precisely why it should never be used to mask uncertainty. The real skill isn’t delivering a thrust; it’s being able to explain why you are or aren’t using one. Confidence built on clarity will always outlast confidence built on theatrics. Patients don’t need us to look impressive; they need us to be right often enough, honest always, and reflective when things don’t go to plan.

I’ve built a career around this exceptional tool, where I know the hands-on skills I’ve delivered, with precision, intent, and certainty, have genuinely helped lead patients out of pain. The change can be immediate. Shoulders drop. Faces soften. Someone stands up straighter than they walked in. Painkillers stopped. In those moments, it’s tempting to believe you’ve found the answer. The only answer, one might think. And sometimes, that confidence becomes contagious, for not only for the patient, but also the practitioner.

I’ve watched patients return with a familiar look: hopeful, expectant, already positioning themselves on the table before I’ve finished asking how they’ve been. “that crack was unreal!”, they’ll say. “That’s what fixed it last time.” Cue the affectionate but dangerous mythology, and the uphill task of convincing them to engage in load management, to change their habits, to take exercise snacks in the day, that will also help them.

That’s when the line becomes blurred. Not because manipulation is ineffective - far from it - but because it worked so well that it risks becoming the whole story rather than part of it. When relief is rapid and repeatable, patients can start chasing the sensation rather than the outcome. And if we’re not careful, clinicians can start supplying it, mistaking patient satisfaction for patient progress.

This is where certainty needs tempering with responsibility. The same confidence that makes manipulation powerful is the same confidence that can quietly undermine long-term resilience if it isn’t paired with education and restraint. Because our job isn’t to create repeat customers for a noise or a feeling, it’s to create people who don’t need us quite so much.

Used well, manipulation can open a door. Used carelessly, it can become a revolving one.

A chiropractor’s duty of care means knowing when not to adjust. That’s right - sometimes the best adjustment is no adjustment at all. Think of it like cooking: just because you own a spice rack doesn’t mean every dish needs a tablespoon of chilli powder. We’re responsible for screening, assessing, and deciding whether manipulation is appropriate, or whether other treatments (rehab, soft tissue therapy, education) might serve the patient better.

Part of the confusion stems from terminology. In chiropractic circles you’ll hear ‘adjustment’, in physio circles ‘manipulation’, and in sports medicine ‘high-velocity low-amplitude thrust’ (HVLA), because nothing says fun like a name that sounds like a rocket launch. Ultimately, the principles are the same: skilled joint mobilisation to restore motion, reduce pain, and improve function. In musculoskeletal (MSK) practice and elite sports, it’s often used as part of a wider plan, not a solo act.

Context matters enormously. In elite sport, for example, manipulation may be used to restore short-term range of motion or reduce protective tone or stimulate muscle proprioception before training or competition, in full knowledge that it’s a temporary window rather than a permanent fix. In primary care, the same technique might be used far more sparingly, prioritising reassurance, load management, and long-term self-efficacy instead. The technique hasn’t changed, the reason for using it has. This distinction is often lost in online debates, where manipulation is discussed as though it exists in a vacuum rather than within a broader clinical ecosystem.

Equally important is recognising the difference between can and should. Many joints are manipulable; far fewer genuinely need it. Clinical maturity often shows itself not in how many techniques you’ve mastered, but in how many you consciously choose not to use. This restraint isn’t a loss of skill, it’s evidence of it.

Now, let’s talk about the ‘gurus’. Many of us were taught that THE adjustment is the ultimate, all-powerful treatment. And yes, done well, it can be transformative. But as a wise man in red and blue Lycra once said: “With great power comes great responsibility.” Spiderman probably wasn’t talking about the lumbosacral junction, but the point stands. If all you have is an adjustment, every problem starts to look like a subluxation, or, as the saying goes, if you’re a hammer, everything is a nail.

Growing up, I had a very simple engineering lesson passed down from my grandfather, Chuck. He worked out of a garage with a handwritten sign taped to the door: “Moving too much? Weld it. Not moving enough? WD-40. Otherwise, leave it alone.” Alongside it sat his personal motto: “If it ain’t broken, don’t fix it.” As crude as it sounds, that algorithm has served me surprisingly well in understanding when and why I use SMT. In my early chiropractic training, there were moments when the internal monologue behind an adjustment felt suspiciously like: “Why am I manipulating this part of the spine? Well… because chiropractic.” It wasn’t until I began applying my grandfather’s logic: intervention only when justified, that manipulation stopped being a default behaviour and started becoming a considered, rational clinical choice.

As clinicians develop, there’s often a quiet shift from technique obsession to outcome obsession. Early on, it’s tempting to chase the perfect setup, the perfect contact, the perfect cavitation or click. Later, the questions change: Did this meaningfully alter the patient’s symptoms? Did it help them move, train, sleep, or cope better? Did it move the plan forward? If the answer is no, then the elegance of the thrust becomes largely irrelevant. Manipulation earns its place not by tradition or identity, but by utility.

This is where reflective practice becomes essential. Reviewing outcomes, retesting, and being willing to say “that didn’t work as hoped” separates evidence-informed clinicians from ritual-based ones. Manipulation should survive scrutiny; if it doesn’t, it should be modified, replaced, or dropped altogether. Diagnostically the outcomes (or lack of) post-manipulation help to tell an important story of your patient.

And that brings us neatly to the idea of ‘bang for buck’. In patient management, it’s about choosing interventions that deliver the greatest benefit relative to effort, risk, and cost. Just because a treatment exists or your own practitioner bias exists, it doesn’t mean it deserves top billing in every plan. Sometimes a well-placed adjustment will provide immediate relief and functional improvement, but other times targeted exercises, posture education, or soft tissue work will give more lasting results for less fuss. A savvy clinician weighs the options, considers the patient’s goals and context, and selects what truly gives the most value; not what simply makes the loudest crack. It’s a reminder that effectiveness is about outcomes, not theatrics - patients deserve more than just one cute thing.

Let’s not leave patient safety out of this article either. A good chiropractor doesn’t just click and hope for the best, they continually evaluate its appropriateness. We explain what we’ve found, rule in or rule out manipulation as appropriate, and give patients clear advice on what to expect, including when to come back, and more importantly, provide the safety net of when to seek urgent medical help. Done properly, this process builds trust and ensures manipulation is used at the right time, for the right reason, in the right person. The wise words of my former boss, Ulrik Sandström, still resonate with me: “It works for some of the people, some of the time”, much like any other therapy, not just spinal manipulation.

Perhaps the most underappreciated role of manipulation is its ability to open a conversation rather than bring it to a close. When used appropriately, it can help build trust, reduce fear, and give patients a sense of momentum early in their care. However, that moment should act as a doorway, not a destination. The real work often happens afterwards — in the exercises patients actually complete, the beliefs they quietly reframe, and the confidence they rebuild in their own bodies.

Too often, this is undermined by what can only be described as a verbal chiropractic beat-down. Patients are told, “Your head is too far forward”, “Your hips are rotated inwards”, “You’ve got the flattest feet I’ve ever seen”, or “This muscle isn’t working — it’s very weak”. Faced with a list of perceived faults, it’s hardly surprising that the default conclusion becomes that spinal manipulation is the only solution. The problem isn’t the intervention itself, but the narrative built around it.

I often say to students when teaching communication skills: if your patient wasn’t body-dysmorphic when they walked in, there’s a real risk they will be by the time they leave. This kind of fear-based messaging is rarely ill-intentioned, but it reinforces a passive model of care. It keeps patients dependent on being ‘fixed’, rather than empowering them through a more active, collaborative, and resilient approach to recovery.

When framed this way, manipulation stops being the headline act and becomes what it arguably always should have been: a supporting character. Useful, sometimes impressive, occasionally indispensable, but never the entire story and a very essential working cog.

In short: manipulation is neither mystical nor malevolent. It’s a tool - a powerful one at that - but not the only one in the box. Our job is to use it wisely, avoid overuse, and remember that sometimes the most heroic thing we can do for a patient is advise, reassure, educate, signpost…or just leave it be.

Neurodiversity is estimated to affect 15% of the population [1], and despite a similar cause and a great deal of crossover, there are a variety of different diagnoses and expression of traits. Although many neurodiverse profiles are spiky by neurotypical standards, stigmatising terms implying severe disability such as ‘idiot savant’ have thankfully fallen out of use. A spiky profile means there are areas where ability is very high but others where natural ability is very low (when compared to a neurotypical normal range; an important caveat). Many neurodiverse individuals struggle with similar aspects of life because they are expected to conform to societal expectations and etiquette based on neurotypical norms, not neurodiverse ones: and the struggle to fit in whilst hiding your authentic self is exhausting and can lead to overwhelm and dysregulated behaviour. [2]

With neurodiversity deriving from a difference in the detection and transmission of sensory input, it’s no surprise that there’s a large crossover with hypermobility. Less ascending information due to less resistant tissues means the gain is turned up to get enough information to coordinate and perform physical function. The brain will listen harder, gaining more information at greater volume but with less specificity than in neurotypicals. Other senses can be used to compensate (e.g. visual feedback for balance and auditory feedback from heel strike), but these make the performance of function much more concentrated and effortful. The hypermobile patient always walks on sand, never pavement. Also, the necessity of having much more input across the board, to get the information you need, means processing and reaction times can be slower as the system takes longer to filter out the relevant from the irrelevant. And with this increase in volume across the board, it’s no wonder the rates of chronic pain are so much higher. [3] If we expand this model of hypermobility to all mental and physical functions relying on sensory input and feedback, you start to get some idea of the day to day struggles neurodivergent people face.

On the plus side, the sacrifice of speed for detail means neurodiverse individuals will often see the world differently and more completely, so it should be no surprise that a higher percentage of workers in innovation and the creative industries identify as neurodiverse. [4]

In 2011 I took a year out to study for an MSc in Loughborough. At a crossroads in my career, I had become bored and a little disillusioned as a clinician and wanted to step into something new for a while.

When I announced my plans to study for an MSc in Ergonomics, my physio colleagues were perplexed as to how studying someone’s desk set up could possibly be stretched out over the course of an academic year, never mind how it might result in being awarded a postgraduate degree!

I wasn’t sure whether to be amused, insulted or frustrated.

I was mildly amused that my colleagues genuinely thought that I was taking a year out to study this limited topic, a little insulted that they thought I really was that boring but perhaps more frustrated at the tendency of my profession to view things through a reductionist lens. But then again, Google seems to agree with my colleagues’ view, as a quick search of the term returns a plethora of information about the dangers of sitting, lifting or even sleeping in the ‘wrong’ posture.

Ergonomics is in fact the science of designing things (jobs, products, systems and environments) to allow people to work safely, efficiently and comfortably. To reduce this whole discipline to ‘that’ diagram demonstrating the ‘correct’ way to sit and stare at a computer screen doesn’t merely oversimplify the field but is incorrect. Most ergonomists will tell you that as dynamic beings designed to move, the best posture is, in fact, the next posture!

The truth is, those of us working within musculoskeletal pain can learn a great deal from Ergonomics, but it has nothing to do with posture or the correct way to lift a box. Far from studying one thing in detail, embarking on this MSc taught me about a range of topics, but more than that, it shifted my thinking and altered my career path. But more of that later.

The course initially appealed to me because there was a module entitled ‘Product Design’ and I thought that it might allow me to combine my knowledge of human anatomy and function with an ambition to do something a little more creative. On that front, I was to be disappointed. Even though the course was based in the Design school, and Ergonomics certainly underpins the design and development of many products, development of which demands robust research processes, a grasp of statistics is perhaps more valuable than creative flair.

But that’s not to say that my year at Loughborough was all data patterns and standard deviations. Before I get into the part of Ergonomics that I really fell in love with - the part that changed my thinking, and my career path as a physiotherapist - I would like to mention some of the great stuff that I did as part of my Masters year at Loughborough with the aim of giving some insight into such a varied and interesting speciality; one that is all around us, every day, without us often even knowing.

Environmental Ergonomics gave me the opportunity to work on some fascinating projects spending time testing different objects in the wind tunnels in the labs at the university, as well as travelling to Portsmouth to assess sound levels at a rehearsal of the Band of the Royal Marines. Temperature, airflow, noise levels: those are all ergonomic factors.

We’ve all heard the following words before and if we’re honest, probably even said them ourselves:

“Don’t worry, I’ll do my exercises at home.”“I’ve got an incredible idea for a new venture.”“I really need to spend more time on the business, not just in it.”

They’re statements of optimism and enthusiasm, small bursts of conviction that feel productive. But more often than not, they don’t lead to the outcomes we expect.

We can all relate to the patients who fall behind on their rehab, leaders who struggle to make impact, or aspiring entrepreneurs with great ideas that stay on the shelf collecting dust.

Why does this occur? Because of the intention–action gap. The space and disconnect between what people intend to do and what actually gets done.

Good intentions are wonderful, but they don’t create results. And while this may sound painfully obvious, it’s one of the most overlooked truths in human performance and business alike. Simplicity is deceptive, especially when it challenges our human behaviour.

In musculoskeletal rehab, leadership, and entrepreneurship, we consistently see people substitute the feeling of intention for the discipline of execution. ‘Busyness’ becomes the narrative, output masquerades as outcome, and genuine progress quietly slips out of reach.

So, what’s the antidote to consider? Over the past year, I’ve continued to focus on a deceptively simple, three-part framework that has clarified how to bridge the gap between good intentions and real results:

1. Clarity of outcome

Be explicit about where you want to end up. What specifically do you want to achieve and, just as importantly, how will you know when you’ve reached it?