Cookie policy notice

By continuing to use this site you agree to our cookies policy below:
Since 26 May 2011, the law now states that cookies on websites can ony be used with your specific consent. Cookies allow us to ensure that you enjoy the best browsing experience.

This site is intended for health professionals only

At the heart of general practice since 1960

What you need to know about the painful shoulder

Orthopaedic surgeon Mr Brian Cohen answers GP Dr David Morris’s questions on painful arc syndrome, diagnosing a rotator cuff tear and the evidence for steroid injections.

Orthopaedic surgeon Mr Brian Cohen answers GP Dr David Morris's questions on painful arc syndrome, diagnosing a rotator cuff tear and the evidence for steroid injections.

1 We frequently see painful arc syndrome in primary care. It's a clinical finding that can arise from a variety of diagnoses, including supraspinatus tendonitis, subacromial bursitis, subacromial impingement and arthritis of the acromioclavicular joint. But how can we distinguish between them?

41242851‘Painful arc' is a clinical description of pain in the mid range of abduction of the shoulder. The pain is derived from the rotator cuff and the subacromial bursa – the synovial membrane that covers it. These structures pass beneath the coracoacromial arch – consisting of the thick band of the coracoacromial ligament and anterior margin of the acromion. The rotator cuff and subacromial bursa can also be impinged by the bone spurs on the under-surface of the acromioclavicular joint.

Clinically they may all produce a similar symptomatic picture on presentation – pain on use of the arm above shoulder height or discomfort in the shoulder at night that may wake the patient.

Examination may show localised swelling and tenderness with acromioclavicular arthritis as well as so-called ‘end-range' pain in the upper range of abduction of the arm beyond 130° and pain on adduction across the body.

The clinical history and examination may only go so far in establishing the specific diagnosis, as there are often no signs or symptoms that can reliably distinguish supraspinatus tendonitis, subacromial bursitis and subacromial impingement.

In fact, there is often coexistent pathology, with supraspinatus pathology leading to rotator cuff weakness and therefore loss of the depressor action of the cuff on the humeral head. This leads to the humeral head moving upward, impinging the cuff beneath the coracoacromial arch and subacromial bursitis.

An ultrasound or an MRI scan is helpful in distinguishing whether subacromial or acromioclavicular pathology is predominant and this is important in directing treatment.

2 I recently saw a 63-year-old man with painful arc syndrome who hadn't improved after two weeks of ibuprofen. Under these circumstances would you think it worthwhile to try an alternative NSAID or would you instead move on to a local steroid injection? And at what stage would you involve the physiotherapist?

Painful arc syndrome in this age group is frequently caused by subacromial impingement and refers to a patient experiencing pain in an arc of movement between about 70° and 120° of abduction or elevation of the arm. I doubt whether a change of anti-inflammatory medication would be of much benefit.

If you are sure that this patient does have a painful arc syndrome a steroid injection into the subacromial space is more likely to be of benefit, although – as

I outline in question 4 – reliable long-term benefit is questionable. I often recommend a delay of a week or two before starting physiotherapy as there can be brief exacerbation of shoulder pain for a few days following a steroid injection (steroid flare). Physiotherapy would be aimed at rotator cuff stretching and strengthening exercises.

3 How would you demonstrate a significant rotator cuff tear and how should a tear – as opposed to a tendonitis – be managed?

Imaging of the rotator cuff using either ultrasound or MRI is the standard means by which a rotator cuff tear is demonstrated.

But the definition of a ‘significant' rotator cuff tear varies – from clinician to clinician and from patient to patient.

It may mean a painful partial or full thickness tear causing pain, weakness and limitation of shoulder movement, or it may, for instance, mean a full thickness tear of a certain size involving one, two or three components of the rotator cuff: supraspinatus, infraspinatus and subscapularis.

A number of studies have shown that rotator cuff disease increases with age and perhaps up to 50% of people in their mid-60s have bilateral asymptomatic rotator cuff tears. Longitudinal studies have shown that tear size tends to increase over time1 and asymptomatic tears may become painful, but the factors that determine this are unclear.

So management depends on a wide range of factors including the age of the patient, functional or sporting aspirations, the severity of symptoms, whether the tear is partial or full thickness, the degree of tendon retraction and the quality of the muscle – defined by the degree of fatty infiltration.

Treatment options are physiotherapy, steroid injection, subacromial decompression,rotator cuff repair or even muscle transfer, and the specific treatment depends on the individual case.

4 What's the evidence to support use of steroid injections for a painful shoulder? What site of injection should be chosen for different conditions and how often can an injection be repeated at the same site?

Although steroid injections in and around the shoulder are said to be of benefit for shoulder pain, an evidence base is lacking2.

Steroid injection is commonly used as treatment for subacromial impingement and rotator cuff disease by injecting into the subacromial space, best located via a posterior approach in my opinion.

Injecting into the glenohumeral joint is used for treatment of adhesive capsulitis and this is best administered using some form of image guidance such as ultrasound, as blind injections are often inaccurate.

The accuracy of injecting a steroid into the subacromial bursa is at best 70% – even in the hands of an experienced shoulder surgeon3,4 – and ultrasound improves accuracy.

A recent study on the efficacy of cortisone injection into the acromioclavicular joint found short-term benefit in 30% of patients5.

These findings are echoed by a meta-analysis of 26 published trials of steroid injections, which found – despite many randomised controlled trials of steroid injections for shoulder pain – their small sample sizes, variable methodological quality and heterogeneity meant that there was little overall evidence to guide treatment2.

Subacromial steroid injection for rotator cuff disease may be beneficial although its effect may be small and not well-maintained.

Similarly, intra-articular steroid injection may be of limited, short-term benefit for adhesive capsulitis.

Other important issues that remain to be clarified include whether efficacy is influenced by the accuracy of needle placement, anatomical site, frequency, dose and type of steroid.

5 During a recent surgery a middle-aged man with symptoms and signs of a rotator cuff tendonitis requested an X-ray. I explained that this was unlikely to be helpful and suggested rest and a course of NSAIDs. Under what circumstances would you to request a shoulder X-ray or consider an MRI scan?

A radiograph of the shoulder in such a patient – with signs and symptoms of rotator cuff tendonitis – is indeed often normal.

But there may be subtle radiographic signs to support a diagnosis of rotator cuff tendonitits, such as cystic change within the greater tuberosity as a sign of rotator cuff disease or sclerosis and remodelling of the inferior acromial surface that correlates with a rotator cuff tear.

For the most part, though, a radiograph would only indirectly be of help in management in that it would exclude other pathology.

Radiographs of the shoulder would be indicated in general practice in a patient:

• with a history of trauma

• with acute onset of severe unremitting shoulder pain such as with an acute calcific tendonitis

• who is elderly – to identify glenohumeral arthritis or secondary degenerative change from long-standing rotator cuff pathology.

The standard views obtained are an anteroposterior view and an axial view. It is important to appreciate that plain radiographs are very often normal in the presence of a full thickness rotator cuff tear. An MRI scan of the shoulder is particularly helpful following trauma as it may reveal an occult fracture or if the patient is suspected of having subacromial impingement with rotator cuff pathology.

6 What advice can you give to help us avoid missing a shoulder dislocation?

It depends whether you mean a dislocation of the glenohumeral joint or a dislocation of the acromioclavicular joint, as both affect the shoulder girdle.

The diagnosis of an acute dislocation of the glenohumeral joint is made on the basis of a history of trauma, often of the arm being forced into abduction and external rotation or of previous dislocation.

Clinical examination reveals flattening of the deltoid contour over the cape of the shoulder and fullness anteriorly with marked restriction of shoulder movement. Radiographs of the shoulder including anteroposterior, axillary or the so-called ‘Wallace view' should be requested and these would confirm an anterior dislocation – the most common type.

If a dislocation of the shoulder is suspected then it is extremely important to document the status of the axillary nerve prior to reduction, by noting whether the sensation over the ‘regimental badge' area of the shoulder is normal.

It's usually not possible to test whether the motor component of the nerve is intact by assessing contraction of deltoid, as the shoulder is usually too painful to move.

A more difficult scenario is the patient in whom the shoulder subluxes and spontaneously reduces as they just present with post-traumatic pain and limited shoulder movement. Specialised radiographs such as a Stryker notch view may be helpful, looking for a hatchet or Hill Sachs lesion of the posterolateral aspect of the humeral head, indicative of a previous dislocation.

Dislocation of the acromioclavicular joint is caused by a fall onto the point of the shoulder and presents as an obvious swelling and deformity at the lateral margin of the clavicle. Plain radiographs confirm the diagnosis.

7 Sometimes the clinical picture suggests that the origin of shoulder pain lies outside the shoulder joint itself. What other diagnoses should we consider as possible cause of referred pain to the shoulder?

Possible causes of referred pain to the shoulder include neck pain that often radiates to the posterior aspect of the shoulder and to the scalp and cervical nerve root impingement, and that may radiate down the arm below the elbow and be associated with sensory changes in the hand.

Pain may be referred to the shoulder from diaphragmatic irritation, classically from gallbladder disease to the right shoulder and from the heart to the left shoulder.

8When would you recommend that shoulder pain be referred to an orthopaedic specialist? Are there red-flag symptoms that we should be on the lookout for as GPs?

I would recommend that patients with severe unremitting shoulder pain not responding to simple analgesia or anti-inflammatory medication – and associated with a restricted range of shoulder movement – should be referred.

Pain at night that disturbs sleep often indicates significant pathology.

Shoulder pain in children is unusual in the absence of trauma and certainly this is a specific group that deserves close attention.

The patient who has a persistently painful shoulder following trauma with normal radiographs should be referred.

I regularly see such patients who will then come to have a diagnosis of an occult fracture of the proximal humerus, acromioclavicular joint injury or a rotator cuff tear.

9 What can now be achieved through laparoscopic shoulder surgery?

Arthroscopic shoulder surgery techniques have developed immeasurably during the past decade and it is now common to treat a wide range of conditions by this method – such as glenohumeral joint stabilisation with labral repair and capsular plication, subacromial decompression and rotator cuff repair, removal of loose bodies, debridement of calcific deposits within the rotator cuff and more recently arthroscopically assisted stabilisation of the acromioclavicular joint.

10 What are the key clinical findings of a frozen shoulder and how can it be distinguished from a tendonitis of the shoulder? What is the evidence-based management of a frozen shoulder?

Primary idiopathic frozen shoulder or adhesive capsulitis is a one of the most common causes of shoulder pain and limitation of shoulder movement. The condition typically affects patients between 40 and 60 years of age and is slightly more common in women than in men. There is no predisposition based on arm dominance or occupation.

It is more common in diabetes, particularly type 1 diabetes, and it is also more severe and resistant to treatment in this group. It is also associated with epilepsy and hypothyroidism.

Frozen shoulder commonly presents with constant severe pain affecting sleep and limitation of shoulder movement. The key clinical findings on examination is of severe restriction of external rotation in the presence of a normal radiograph. It may be distinguished from tendonitis in that limitation of movement is not so severe and rotation is preserved in this condition.

A wide range of treatment options are available, including physiotherapy, distension and steroid injection of the glenohumeral joint either under ultrasound guidance or arthroscopically, long-acting steroid injections, manipulation under anaesthetic and arthroscopic capsular release.

But the evidence base for management is limited. One review of five trials to determine the effectiveness and safety of arthroscopic distension of the glenohumeral joint concluded that there is ‘silver' level evidence that arthrographic distension with saline and steroid provides short-term benefits in pain, range of movement and function in adhesive capsulitis – but it is uncertain whether this is better than alternative interventions2.

Another evidence review analysed the results of steroid injection treatment for adhesive capsulitis in four randomised controlled trials and suggested that suggested that multiple steroid injections improve pain and range of motion in the short term (six to 16 weeks) from the first injection and that there was evidence that up to three injections were beneficial7.

Mr Brian Cohen is a founding partner of The London Orthopaedic Clinic . He is an upper limb surgeon, practises out of 30 Devonshire St, London, and operates at the Princess Grace Hospital

The London Orthopaedic Clinic holds free monthly education sessions aimed at GPs, physiotherapists and other allied health professionals. For more information visit the website or call 020 7186 1000 for details of the education programme and the clinic's after-hours emergency service

thps shoulder What I will do now What I will do now

Dr David Morris considers the responses to his questions

It's useful to read how judicious use of ultrasound and MRI scans may help achieve a specific diagnosis and direct treatment, so I'll have a lower threshold for referring patients with persistent shoulder pain to orthopaedics.
It's useful to be reminded that shoulder pain may arise from neck pathology and that this possibility should be covered in the history and examination.
If anatomical landmarks are difficult to locate or if a previous steroid injection has been unsuccessful, I'm now going to be more likely to refer to an orthopaedic service where injections can be guided by ultrasound.
But it's also interesting to read that
a relatively common treatment has such a poor evidence base.
In my experience some patients recoil at the thought of shoulder surgery. Being able to tell them about the increasing range of minimally invasive laparoscopic procedures may encourage them to take advantage of the benefits available from surgery.

Dr David Morris is a GP in Shrewsbury, Shropshire

A rotator cuff tear (in red) is best diagnosed by MRI (pictured) or ultrasound MRI scan of a rotator cuff tear

Rate this article 

Click to rate

  • 1 star out of 5
  • 2 stars out of 5
  • 3 stars out of 5
  • 4 stars out of 5
  • 5 stars out of 5

0 out of 5 stars

Have your say