Assessing Shoulder Pain Part 2: no shells, just pearls.

What tools do we have to guide treatment and prognosis with our shoulder pain patients?

Shoulder special tests, even when we know the psychometrics, aren’t doing much to guide our treatment, nor are they helping us to prognosticate.

So where does that leave us?

The history is an obvious start. The patient, with some prompting, will reveal his irritability, his aggravating factors, his easing factors, and you can use this to give him solid advice on activities to avoid and perform. You’ll also get an idea of some of the psychosocial variables that are at play. Fear-avoidance, catastrophizing, competing motivations, depression, anxiety, distaste for work, etc. This will guide your words, which is likely the most powerful treatment tool in our box.

But that stuff, as crucial as it is, is well worn. Let’s discuss a few other pragmatic ideas that will give us tangible information to guide treatment.

Idea #1: Impairment Hunting. Find things that are limited that you can affect with treatment. For instance, decreased internal rotation? (okay, I can treat that), decreased thoracic rotation? (hey, I can treat that too), weak rotators? (I know how to facilitate and strengthen those). The list goes on.

Ear mark these impairments to apply an intervention to. But how do you know if these impairments are relevant to their ‘issue’? Enter Asterisks Signs!

Idea #2: Asterisks Signs! An Asterisks Sign is otherwise known as a concordant sign or sometimes as a test-retest. It’s a movement that reproduces their familiar symptoms. Ideally you will have a few of them that are quick and easy to perform. It’s also useful to have one that is a meaningful movement to the patient. The idea is to have a quick movement to do before and after an intervention, which will immediately tell you if your intervention made a positive difference.

This means assessment doesn’t stop once treatment begins. It also means that you will never again waste more than about 30 seconds on an inappropriate treatment. Another advantage of asteriks signs is that you’re demonstrating treatment effectiveness to your patient. Your patients will quickly buy in to your treatment plan when they observe obvious changes that you’re making in real time.

On a personal note, I find that Asteriks Signs make clinical work a lot less stressful. Before using them, I’d work off a theoretical idea based on weak or invalid patho-anatomical assumptions. All I could then do was cross my fingers for my patient’s symptoms to get on board with my theory. With an asterisks sign to test against, there’s no more wishing and hoping, and I don’t need to believe in a biomedical idea of pathological anatomy that again and again doesn’t hold up in practice or in research.

Idea #3: Symptom modification. I reckon that this came out of Brian Mulligans work. The idea is to facilitate a movement by way of a glide, postural shift, or facilitation to see if that reduces or eliminates their asterisks sign. This can easily be done during the assessment. For example, does their pain free range improve if you facilitate posterior tilt and upward rotation of the scapula? What if you simply get them to extend in their thoracic spine? Another one to try: What if they do a 45 second bicep isometric and that improves their asterisks sign? Maybe that points to an analgesic effect on a biceps tendinopathy (ref). In any event, orienting an assessment towards symptom modification gives you great direction on how to modify their pain.

Idea #4: Treatment Based Classification (TBC) This is the ultimate pragmatic shift in assessment. Basically, we go from diagnosing them with a condition, to categorizing them into a subgroup that responds well to a certain collage of treatments. You categorize them based on predictive variables. For example, in the low back, if your pain has been present for <16 days and you don’t have pain below the knee you’re patient would fall into the manipulation subgroup. So you’re not calling it a disc, a facet joint, or even a schmorels node; instead, you’re ‘diagnosing’ patients into a specific group that will do well with a specific treatment. Now that is user friendly. And bonus, it’s evidence based (ref)!

Schellingerhout and colleagues, in 2008, called for the development of a treatment based classification for shoulder pain after observing that there wasn’t uniformity in the diagnostic labels that were being used (frozen shoulder, impingement, instability, etc). They also found that there was categorically weak inter-observer reliability at coming to the same diagnostic label for patients with shoulder pain. The systematic review also found that treatment efforts based on these diagnostic labels seemed to all fail. This reaffirms last weeks post, in that we need to move on and away from special tests and the patho-anatomy that they espouse.

Problem is that the research hasn’t caught up to this good idea. There are a few sub group proposals out there but they’re nowhere near validation.

For now, prevailing advice is to categorize based on a couple straightforward questions:

  1. Is this shoulder generally tight or loose?
  2. Is this shoulder generally weak or strong?

Let’s recap before getting into prognosis:

Well first off, we don’t care so much about pain generating structures. Secondly, we have an idea of what impairments they have, what makes them squeak (asterisks signs), what modifies their symptoms, and if they’re generally tight, loose, weak, or strong. Furthermore, from the history we know about their severity and irritability. This adds up to a great deal of information; and importantly it is all relevant and useful for guiding treatment.

But what about prognosis?

3 epidemiological findings were much more effective at predicting chronicity: 1) high pain intensity, 2) concomitant neck pain, and 3) longer duration of symptoms.

Well okay, some special tests are pretty good at telling you about instability (apprehension-relocation, load & shift, etc), others are pretty good at telling you there’s a complete tear (bear hug, hornblowers, lift off, drop arm in the Park cluster), and this may impact your projection about how long it will take for your patient to get better. But there’s not much literature out there about how these findings actually affect prognosis. 

On that front, a fellow by the name Kooijman and colleagues looked at prognosticating shoulders and found that 3 epidemiological findings were much more effective at predicting chronicity: 1) high pain intensity, 2) concomitant neck pain, and 3) longer duration of symptoms. These three findings, all easily obtained in the interview, were the only consistent variables associated with persistent symptoms in a primary care setting. That’s a lot more straightforward and user friendly at guiding your prognosis than basing it on patho-anatomy.

So the moral of the story: be pragmatic with your assessment, don’t get lured into the swamps of biomechanics and patho-anatomy, and stay on point with asterisks signs. Keep physio fun!

Check out and register for our upcoming course with Adam Meakins: The Shoulder: Complex doesn’t have to be complicated.