STAR Shoulder Pain Classification System

There's currently a hot debate going on about how we should categorize, diagnose, and treat shoulder pain. Classic diagnoses like impingement have been sharply criticized, so Austin takes us through an alternative system that may serve you well in the clinic and on the OCS: the Staged Approach to Rehabilitation Classification: Shoulder Disorders, or STAR-Shoulder approach, proposed by McClure and Michener in PTJ in 2015.

Use code FIELDGUIDE for 40% off a MedBridge subscription.

Find more resources and subscribe to practice questions at PhysioFieldGuide.com.

Transcript

Hello and welcome back to the OCS field guide podcast. Before we start, I want to tell you about our 2021 Christmas giveaway in case you haven’t seen our posts yet. We want to say thank you to all of our supporters that have made this podcast possible and get the word out to other busy clinicians in your circles who may be considering taking the OCS or just want an easy way to refresh on evidence. You’ll have a chance to win either AirPod Pros, or a $100 Amazon gift card. So head over to our Instagram @OCSfieldguide. All you have to do is follow our account, make a comment on the giveaway post that tag another PT friend that would be interested in our content, and share the post to your story, mentioning your favorite episode. Winners will be picked shortly after the cinnamon rolls are eaten on Christmas morning, so be sure you are entered by then.

Let’s get on with it, in this episode I’ll cover what appears to be the prevailing shoulder pain rehab classification system as put forth by McClure and Michener in the “Staged Approach to Rehabilitation Classification: Shoulder Disorders,” or “STAR-Shoulder” article, published in PTJ in 2015.

This podcast is not going to be packed with facts that you need to memorize. Instead, I’m covering this because it provides an excellent framework for the critical thinking necessary to make good decisions when it comes to shoulder cases on the exam. The shoulder is very messy if you look at it only from a pathoanatomic model, but the STAR-shoulder article does a good job of respecting pathoanatomic diagnosis, while highlighting the importance of treatment based on individual impairments and level of irritability. Pathoanatomic diagnosis does inform the therapist somewhat about expected tissue-healing, pathology, prognosis, depending on the diagnosis, and depending on if a specific diagnosis can be made, and thus it should still inform treatment, but it does not tell you much about the impairments. Two patients with the same diagnosis may have very different impairments and levels of irritability and thus need very different treatment, and two patients with different pathoanatomic diagnoses could have very similar impairments and irritability and thus need very similar treatment. A single patient will typically have the same pathoanatomic diagnosis throughout treatment, but their impairments and irritability will change and develop, and so should treatment. The article gives two good examples for the need to blent pathoanatomic diagnosis, impairments, and irritability for best management. Someone having the diagnosis of adhesive capsulitis does tell you to expect the patient to have loss of ROM and protracted course of treatment, but it does not tell you anything about what motions are most limited, or how irritable the condition is. Likewise, knowing that a patient has sustained a bankart lesion is very important in that it will guide the therapist to appropriately limit early exposure to external rotation ROM, but will not inform the therapist about the specific impairments that will guide the bulk of the treatment for that individual. Additionally, as we’ve learned more about the limits of imaging and special tests, we often cannot be certain that a patient has one specific pathoanatomic diagnosis over another, or that a finding on imaging is the cause or source of their complaint. Thus the STAR-shoulder classification system attempts to group together pathoanatomic diagnoses that are most likely to have similar presentation and special considerations that will affect treatment, but treatment is selected based on an integration of knowledge of pathoanatomic diagnosis when pertinent, individual impairments, and level of irritability. 

Let’s jump right in. This uses a similar staged examination and intervention paradigm to what you have seen in the CPGs. Stage one, screening, involves history, general physical examination, and screening for any red or yellow flags, and determining whether a patient is appropriate for PT, referral and PT, or referral only. They include a good list of the most common serious pathologies that could present as shoulder pain and their red flag history and exam findings so I’ll run through those as follows: tumor, which could come along with history of cancer, signs and symptoms of cancer such as unexplained weight loss, pain not correlated with mechanical stress, and unexplained fatigue, and unexplained mass, swelling, or deformity. Infection, which may have signs or symptoms including redness of the skin, fever, and being systemically unwell. Fracture or unreduced dislocation, which should have history including significant trauma or seizure and present with acute disabling pain, acute loss of motion, and deformity or loss of normal contour. Neurologic lesion, for which you’d look out for unexplained sensory or motor deficit. Finally, visceral pathology, which could come along with pain not reproduced with shoulder mechanical stress, pain or symptoms with physical exertion or respiratory stress, pain associated with gastrointestinal symptoms, and/or scapular pain with ingestion of fatty foods. In fact, a woman I used to work with previously had a couple years of terrible shoulder pain and was bounced around between doctors and PTs and was diagnosed with everything from rotator cuff tear to nerve damage to cervical radiculopathy when really she had a bad gallbladder and a rare liver disease. Stage 2 is actually titled patho-anatomic diagnosis, and involves a more specific examination and determination of pathoanatomic diagnosis, ruling in or out whether the pain is in fact coming from the shoulder, and what diagnostic category the patient best fits in. Unlike the lumbar spine, in the shoulder there are plenty of cases where we can determine a pretty specific pathoanatomic diagnosis and where it does matter such as adhesive capsulitis, AC joint sprain, anterior instability, etc. but for the most common presentation of shoulder pain which we will call subacromial pain syndrome, correlation of patient symptoms to specific pathoanatomic diagnosis is hotly debated, less possible, and less important. The three categories McClure and Michener use are: Subacromial pain syndrome, adhesive capsulitis, glenohumeral instability, and then a fourth “other” category. We’ll dig into those in a second. Stage 3 is placing the patient in a rehab classification based on 1) tissue irritability and 2) their specific impairments, both of which then guide the intervention used. Let’s dig into each category. 

The first, and messiest category, is subacromial pain syndrome. This label is one attempt to get away from the misleading “impingement” label. However, the name the orthopedic section appears to be going with for their planned CPGs is Rotator Cuff Syndrome, and in ICF language it would be “Shoulder pain and muscle power deficits”. Ultimately all three of those titles aren’t great in my opinion, as not all pain in this category has to be from a subacromial structure, not all pain has to do with the rotator cuff, and not all individuals in this category have to have muscle power deficit. But semantics aside, this category includes the presentations we would historically have called subacromial impingement and much more. I like this quote from the article, “The category of “subacromial pain syndrome” is particularly challenging and includes common pathoanatomic labels such as subacromial impingement, bicipital tendinopathy, rotator cuff tendinopathy and tears, subacromial bursitis, secondary instability, and SLAP lesions. The current use of such a large number of pathoanatomic diagnostic categories that are not easily differentiated by a physical examination is impractical and likely does not facilitate treatment decision making for rehabilitation.” That being said, I don’t think you are off the hook yet for knowing things like primary vs secondary impingement, acromion types, SLAP tear types, etc. but that is information for another time. They use the following signs and symptoms to “rule-in” to this category: Impingement signs (such as positive neers, hawkins-kennedy, and jobe or “empty can” test), painful arc, painful resisted testing of rotator cuff muscles, weakness, and in the case of a significant rotator cuff tear, atrophy. You would “rule-out” this category with any significant loss of passive range of motion (key word being “significant” as many of these individuals have smaller but still important loss of ROM in certain directions), or signs of glenohumeral instability. It’s good to remember that “impingement” tests are just tissue provocation tests. They tell you that something in the space you are closing down is sensitive, which is good information, but it does not give a diagnosis or tell you what you need to do to make either not be so sensitive, or not get pinched on so much. Remember that a painful arc or impingement tests are only valid when the history matches up. I guarantee a hawkins kennedy is going to be very positive in someone with adhesive capsulitis, but it does not mean they fit in the subacromial pain category. Now, this classification system does lump in rotator cuff tears in this group, which in most cases is absolutely appropriate.  As research shows compelling evidence for treating even full thickness rotator cuff tears with physical therapy. In fact, the MOON trial (see Kuhn et al 2013 in the journal of shoulder and elbow surgery) showed that 75% of individuals seeking treatment with an atraumatic full thickness have a successful outcome with physical therapy alone and elect not to have surgery. Now this study did exclude acute or traumatic onset rotator cuff tears. However a 2020 RCT in the journal of shoulder and elbow surgery found no significant difference in outcome between surgical repair and physical therapy for the treatment of small acute traumatic rotator cuff tears. The 2019 Cochrane review on rotator cuff surgery for small degenerative tears showed and I quote, “little or no clinically important benefits with respect to pain, function, overall quality of life or participant-rated global assessment of treatment success when compared with non-operative treatment. Surgery may not improve shoulder pain or function compared with exercises, with or without glucocorticoid injections.” So yes, do treat the rotator cuff tear, especially if it’s a degenerative tear, and probably even if it is an acute or traumatic tear, especially if it is small, but try it even if it is bigger. The article does make the appropriate disclaimer that especially larger acute or traumatic rotator cuff tears may be well managed with surgery. Still, I imagine future iterations may consider substantial rotator cuff tears as their own subgroup as evidence and guidance grows for non-operative management.

The second category is adhesive capsulitis, or in ICF terms, shoulder pain with mobility deficits. We’ve already had a podcast going in depth on this diagnosis, so we won’t waste time on it beyond saying you will rule in with insidious onset of progressive pain and ROM loss, especially in individuals between 40 and 60 years old, where active and passive ranges of motion are similarly impaired, with ER being most limited. You’d rule this out in individuals younger than 40, and anyone with normal active and passive ROM. Again, for more depth, check out my podcasts on adhesive capsulitis.

The third category is glenohumeral instability, or in ICF language, shoulder stability and movement coordination impairments. This category includes both traumatic and atraumatic, multidirectional instability and will rule in with age below 40, a history of dislocation or subluxation, positive apprehension test, positive relocation test, and generalized laxity such as with a beighton score of 5 or more out of 9. Of note, the anterior apprehension test is most sensitive and specific when apprehension rather than just pain is judged as positive. To get the most diagnostic test (with a sensitivity of 81% and specificity of 98%) you combine the apprehension test with the relocation test where you place a posterior directed force on the head of the humerus and repeat the test, which should abolish the apprehension in a positive test. Although this article does not include posterior instability signs or tests in the rule-in criteria, but I’d venture a guess those would fit well in this category. Posterior instability would likely have a history of subluxation or dislocation due to a longitudinal posterior directed force on the humerus especially in some horizontal adduction and internal rotation such as blocking a 300-pound lineman, or as our most recent case in the clinic had, landing and mountain bike jump really hard with the force driving through the arm. These individuals would potentially have a positive posterior apprehension test. This test has pretty low sensitivity, but pretty good specificity. AKA you aren’t going to rule out posterior instability with a negative test if they have the appropriate history and mechanism, but a positive rules it in pretty confidently. In case you’re rusty, this test is performed with the patient in supine, by blocking the scapula against the thoracic cage with one hand and horizontally aDDucting the the humerus across the body while applying a posterior force through the humerus. Reproduction of apprehension is considered a positive test. Of course, you would rule out the glenohumeral instability category if there is no history of subluxation or dislocation and if there is no apprehension with testing. This is a good time to note that there, as always, may be cases where people meet criteria for multiple categories, such as the individual with secondary impingement related to glenohumeral instability. They may meet criteria for both, which highlights the importance of specific impairment based treatment. 

Then we have “other”, which could technically be a massive category, but they include the following: post-operative shoulder management, glenohumeral arthritis, fractures, acromioclavicular joint pathology, neural entrapment, myofascial pain, and fibromyalgia. 

No matter whether a patient matches up with one of the three larger categories or has one of the many other pathoanatomic diagnoses of the shoulder, the rehabilitation classification in part three will likely help guide treatment. 

Let’s move on to part 3, rehab classification based on stage of irritability and specific impairments. Similar to what we have seen in the clinical practice guidelines there are high, moderate, and low irritability categories. High irritability is defined as high pain, such as seven or greater out of 10, consistent night or rest pain, pain before end of range of motion, active range of motion less than passive range of motion, and high disability. Interventions here should be focused on minimizing physical stress via activity modification or other means, and we could probably include things like addressing any psychosocial factors or pain education as needed. Moderate irritability is characterized by moderate pain such as 4 to 6 out of 10, intermittent night or resting pain, pain at end of range of motion, active range of motion similar to PROM, and moderate disability. Interventions here will focus on applying mild-moderate physical stress, addressing impairments, and allowing basic-level functional activity restoration. Low irritability is characterized by low pain, 3 or less out of 10, absent night or rest pain, minimal pain with overpressure, and active range of motion equal to passive range of motion. Interventions well progress to moderate to high physical stress addressing impairments and progressing to high-demand functional activity restoration. 

Now for impairments:

The first two impairments are more of a general nature. First, Pain associated with local tissues pathology: for high irritability appropriate interventions include activity modification, manual therapy, and modalities; for moderate, activity modification with progressive activity re-introduction, manual therapy as needed, limited modality use; and for low, they just say no modalities, but I think we could put progressive reintroduction of activity and tissue loading. The other more general impairment is pain associated with central sensitization, and for all levels of irritability they suggest progressive exposure to activity and appropriate medical management. I think had it been written 5 years later, it would be including things like pain neuroscience education, isometrics, general exercise and aerobic activity, and other interventions addressing things like fear avoidance, pain catastrophizing, and pain self-efficacy. 

Now for the more specific impairments. Limited passive mobility, such as of the joint, muscle, or neural structures. The primary treatments are ROM, stretching, and manual therapy across irritability levels, but with progression in degree, intensity, and duration. For high irritability, ROM, stretching, and manual therapy only in pain-free range typically short of end range. For moderate, progressing to comfortable end-range stretching, but intermittent shorter duration. For low irritability, progressing to tolerable end range stretching with longer duration and higher frequency. 

Next is Neuromuscular weakness associated with atrophy, disuse, and deconditioning. For this impairment, with high irritability interventions will focus on active range of motion within pain-free ranges; for moderate will include light or moderate resistance to fatigue through mid ranges; and for low irritability will include moderate or high resistance activity to fatigue, including to end-ranges of motion. 

Then we have neuromuscular weakness associated with poor motor control or neural activation. For high irritability, active range of motion within pain-free ranges is included again and they say to consider use of biofeedback, neuromuscular electrical stimulation, or other activation strategies. For moderate irritability, the focus would be basic movement training with emphasis of quality and precision rather than resistance according to motor learning principles. And with low irritability, interventions focus on high-demand movement training, again with emphasis on quality rather than resistance according to motor learning principles. 

Next is functional activity intolerance. With high irritability cases, intervention will focus on offloading and activity modification to protect the joint or tissue from end range or high load and will encourage the use of unaffected regions; moderate will focus on progressive engagement in basic functional activity; and low will focus on progressively engage in high-demand functional activity. 

Finally, they include poor patient understanding leading to inappropriate activity or avoidance of activity, for which at all irritability levels the intervention is appropriate patient education.

I want to include a couple other impairments not in the STAR article, negative patient expectation and low pain-self efficacy. A pair of high powered studies by Chester et al in 2018 and ‘19 in BJSM studied 34 different potential prognostic factors in individuals with shoulder pain (including things like BMI, smoking status, anxiety, time off of work, sex, duration of symptoms, clinical exam findings, etc), the only two factors that influenced outcome with PT, other than pain severity and level of disability (which we expect), were patient expectation and pain self-efficacy. Those that believed they would have complete recovery vs slight improvement as a result of physical therapy had significantly better outcome and those that had higher pain self-efficacy scores also had better outcomes. The second study even found that those with high pain and disability, but positive expectation and high pain self-efficacy did as well as those who had low pain and disability but low expectation and low pain self-efficacy. So, subjective examination should look for these factors and consider them as high priority impairments to be addressed with education, strategies to achieve buy-in, and appropriate referral as needed. This is so important because I can definitely see the OCS presenting a subacromial pain syndrome case where the patient has been diagnosed with a small full thickness supraspinatus tear with some no-brainer impairments to address like weak external rotators and decreased internal rotation range of motion, but where the patient mentions that they are just doing PT in order to fulfill insurance requirements for surgery. In this case, if it’s available, I would be choosing an option that includes something like “education that physical therapy treatment results in successful outcome for most patients.”

Before we finish I want to briefly give one treatment evidence update. In 2020 JOSPT published an updated systematic review of the effectiveness of interventions for the treatment of subacromial shoulder pain, definitely an important paper for any OCS-takers. I’ll read you the concise conclusion, which should give you what you need to know: Evidence for exercise as the most important management strategy for SSP is increasing and strengthening. Ongoing research is necessary to identify whether there is an optimal dose and type of exercise. Currently, it is not possible to state that one exercise program is more appropriate than another. However, a strong recommendation may be made to include manual therapy as an adjunct intervention with exercise. Conflicting evidence surrounds the effectiveness of multimodal therapy and corticosteroid injection. Other commonly prescribed nonsurgical interventions, such as ultrasound, low-level laser, and extracorporeal shockwave therapy, lack evidence of effectiveness. 

I think the take-away for you is that exercise with the addition of manual therapy, especially early on, is the best treatment for subacromial pain. But since there is no consensus on the best exercise program, you should use a common-sense, impairment-based, irritability-guided, pathoanatomically-respecting treatment strategy put forth in the STAR-shoulder article.

That’s it for today, if you are listening to this episode before Christmas of 2021, be sure to head over to our instagram page @OCSfieldguide to enter to win AirPod Pros or a $100 Amazon gift card.