A letter to a freshly graduated therapist (part 2 – Classification models)

Now that you know the basics, we can get slightly more specific with classification models. These are classifications used in physiotherapy. Keep in mind that whatever classification you use, it should help you refer when needed. We have to understand that we can not treat everything.

I am biased. I believe that classifying patients leads to better outcome and that it is the future of physiotherapy. Hopefully, I will convince you to think the same way.


In university, I was taught to direct my treatment based on the physiotherapy diagnosis. It felt incredible to believe that I was able to diagnose! It felt incredible to think that I was able to differentiate between a bursitis, a rotator cuff tear, a herniated disc, a hypomobile joint and an instability! Each diagnosis had a slightly different treatment! A herniated disc likes traction. A bursitis benefits from ultrasound. A rotator cuff you check motor control of the scapula and drop some deep tendon friction to realign the fibers. A hypomobile joint you give ROM and throw some massage. It felt great! Too bad a few weeks after my graduation, I realized that physiotherapists weren’t reliable at giving a pathonanatomical diagnostic. This leads to poor evidence regarding effectiveness of physiotherapy. It frustrated me. Take for example low back pain. Even after thousands of randomized control trials, there is still no consensus on the best diagnosis and/or treatment!

Let me explain. The tools we use to diagnose aren’t reliable:

Let’s push it further. Let us assume that those tests were reliable… Let us assume that somehow we were able to identify that someone has a disc herniation or an instability. How do we even make sure that it is the cause of their pain? To what do we compare them to? MRI? X-Rays? We can’t.

Here is the reason why:

VOMIT Poster 2014 SAMPLE

Most asymptomatic people (never even had pain in their life) present with abnormalities! This leads to the question: “Is their pain really due to the herniated disc? Or was the herniated disc already present?”

So… Most of the time, we can’t give a patho-anatomical diagnosis, it does not guide our treatment, and it can even lead to wrong treatment approaches.

This is why many researchers got interested in classification models.

Preliminary researches show that classifying patients leads to better outcome (and yes, I cherry pick my studies and I am biased. Make your own opinion):

Mechanical Diagnosis and Therapy (Low back, knee)

Jeremy Lewis (shoulder)

O’Sullivan classification (low back)

Treatment based classification (neck, low back)

Mechanism-based classification (just theoretically)

Theory: http://www.manualtherapyjournal.com/article/S1356-689X(12)00081-1/abstract  . And theoretical treatment matched to the centrally sensitized subgroup: Pain education helps

I believe that more research is needed but that we have enough evidence to start classifying our patients into homogenous subgroups based on their clinical presentation. Each of them would benefit from a specific treatment…We also have enough evidence to stop worrying as much as before about pathoanatomy.

I believe that classifying patients before we treat them will lead to better outcome. This will then (finally) result in more conclusive researches supporting physiotherapy treatments.

I hope this helps.

Olivier Lam


4 Replies to “A letter to a freshly graduated therapist (part 2 – Classification models)”

  1. Impressive content over here Olivier. Perhaps your frustrations over our way of categorizing and treating patients could be solved by looking at what the “Integrated Systems Model”, “Connect Therapy” and the “Thoracic Ring Approach” have to offer. The former are frameworks that looks for every individual patients the SOURCE of dysfunction, the DRIVER for symptoms that can be elsewhere in the body.
    J’en ai fait un résumé et tu peux voir la liste de référence à la fin de mon article.


    1. Thanks for sharing,

      from the little I have read on Diane Lee and LJ Lee work, I can say that it is very similar to other approaches… (symptom modification procedures which tells them what the “source” or “driver” of pain is)

      I personally don’t use it because I don’t know enough about it and I haven’t got time to read/ find any research on it :O

      I will look into it !


    1. Yes, a diagnosis based on pathoanatomy could be seen as a classification based on patient’s pathology. However, we know that it is extremely difficult to give a diagnostic in more than 85% of the cases and that the tests used in physio aren’t reliable/ specific/ sensitive enough to do so. Therefore, giving a treatment based on the pathoanatomical diagnosis is not efficient in most cases.

      When we talk about classification models, we aim to classify patients based on their clinical characteristics (ex: repeated movements that diminish their pain, repositioning of scapula decrease the pain/ decrease ROM/ low fear avoidance beliefs/ etc..)

      Thanks for the comment,
      I hope this makes it a bit clearer!


Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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