This site is intended for Healthcare Professionals only
Back to CKS

Musculoskeletal

Shoulder pain

Last revised in November 2022

Causes of pain arising from the shoulder include:Rotator cuff disorders.Frozen shoulder.Instability disorders.

Shoulder pain: Summary

  • Causes of pain arising from the shoulder include:
    • Rotator cuff disorders.
    • Frozen shoulder.
    • Instability disorders.
    • Acromioclavicular joint disorders.
    • Glenohumeral joint osteoarthritis.
    • Inflammatory arthritis.
    • Septic arthritis.
  • Causes of pain which arise from elsewhere include:
    • Malignancy.
    • Referred pain from the neck, heart, or lungs. 
    • Polymyalgia rheumatica.
  • Assessment of a person with shoulder pain involves taking a history and performing an examination.
    • History should include: pain characteristics, stiffness, instability, functional impairment, trauma, other joint symptoms, neck pain, systemic or neurological symptoms, occupation and sporting activities, medication and past history of musculoskeletal problems, malignancy, or significant co-morbidities.
    • Shoulder examination should include inspection, palpation, and active and passive movements. The neck, arms, axilla, and chest should also be examined for possible referred causes, and neurological examination performed if appropriate. 
  • Investigations are not usually indicated as part of a primary care assessment of shoulder pain.
    • Blood tests should be performed if malignancy, polymyalgia rheumatica, or inflammatory arthritis are suspected. Testing for diabetes should be considered for people with frozen shoulder.
    • X-rays should be considered if there is a history of trauma; there is little improvement with conservative treatment; symptoms last more than four weeks; there is severe pain or restriction of movement; or arthritis is suspected.
  • If any shoulder red flags are identified in the history or on examination, the person should be referred urgently to secondary care. Red flags include:
    • Trauma, pain and weakness, or sudden loss of ability to actively raise the arm (with or without trauma): suspect acute rotator cuff tear.
    • Any shoulder mass or swelling: suspect malignancy.
    • Red skin, painful joint, fever, or the person is systemically unwell: suspect septic arthritis.
    • Trauma leading to loss of rotation and abnormal shape: possible shoulder dislocation.
    • New symptoms of inflammation in several joints: suspect inflammatory arthritis.
  • Emergency orthopaedics assessment should be arranged for people following acute trauma, depending on clinical judgement, and for suspected:
    • Septic arthritis.
    • Shoulder dislocation.
  • Initial management of a person with shoulder pain should include:
    • Explanation of the diagnosis and advice regarding work and activity.
    • The prescription of analgesia, if appropriate.
    • Consideration of physiotherapy referral.
  • A corticosteroid injection may be considered for a person with shoulder pain, depending on the suspected cause and the severity of symptoms.
  • Referral to a specialised musculoskeletal clinic could be considered, for example for provision of physiotherapy. However, if orthopaedic referral is indicated, this should not be delayed. 
  • Referral to secondary care should be considered if pain and function are not improving following conservative treatment for 3 months.
  • An earlier referral should be considered if:
    • The pain is having significant impact on the person.
    • There is recurrent shoulder instability.
    • There is severe post-traumatic pain.

Have I got the right topic?

From age 16 years onwards.

This CKS topic covers the diagnosis and management of shoulder pain in primary care. It covers the management of rotator cuff disorders, frozen shoulder, instability disorders, acromioclavicular disorders, and glenohumeral osteoarthritis. 

This CKS topic does not cover exercise regimens, injection techniques, or surgical management. It also does not cover the management of shoulder pain referred from elsewhere or the management of inflammatory arthritis, Paget's disease, avascular necrosis of the humeral head, or septic arthritis.

There are separate CKS topics on Neck pain - acute torticollis, Neck pain - cervical radiculopathy, Neck pain - non-specific, Neck pain - whiplash injury, NSAIDs - prescribing issues, Osteoarthritis, and Rheumatoid arthritis.

The target audience for this CKS topic is healthcare professionals working within the NHS in the UK, and providing first contact or primary healthcare.

How up-to-date is this topic?

Changes

November 2022 — reviewed. A literature search was conducted in October 2022 to identify evidence-based guidelines, UK policy, systematic reviews, and key RCTs published since the last revision of the topic. No major changes to recommendations have been made. 

Previous changes

January to April 2017 — reviewed. A literature search was conducted in January 2016 to identify evidence-based guidelines, UK policy, systematic reviews, and key RCTs published since the last revision of the topic. Changes include:

  • The differential diagnoses for shoulder pain has been revised.
  • The diagnosis section, 'How do I assess someone with shoulder pain?', has been added. The clinical assessment of someone with shoulder pain has been revised. The Hawkins test has been taken out of the examination, as this is not recommended in the Oxford University hospitals guideline on the diagnosis of shoulder problems in primary care. 
  • A section regarding the initial management of a person with shoulder pain has been added. The criteria for emergency and urgent referral to secondary care have been revised.
  • The management sections for rotator cuff disorders, frozen shoulder, instability disorder, acromioclavicular joint disorders and glenohumeral osteoarthritis have been revised. 
  • For the management of rotator cuff disorder, the following was added:
    • Do not give more than two corticosteroid injections. Normally only one injection should be given (due to the risk of tendon damage from repeated injections). A second injection may be given after 6 weeks, but only to people who have seen improvement following the first injection, and who need the pain relief to allow them to undergo physiotherapy.
  • For the management of frozen shoulder,  the recommendation to refer to physiotherapy if a person can tolerate movement of the affected shoulder was removed. Many physiotherapists perform steroid injections, which can be performed in the pain predominant phase.  
  • For the management of rotator cuff disorders, the recommendation not to give a corticosteroid injection if a rotator cuff tear is suspected was removed. Steroid injections are a management option for degenerative rotator cuff tears. Suspected traumatic rotator cuff tears should be referred urgently to orthopaedics.

April 2015 — minor update. Links have been inserted to the CKS topic on Analgesia - mild-to-moderate pain.

February 2013 — minor update. The 2013 QIPP options for local implementation have been added to this topic.

October 2012 — reviewed. A literature search was conducted in September 2012 to identify evidence-based guidelines, UK policy, systematic reviews, and key RCTs published since the last revision of the topic. A change has been to the recommendation on treatment of rotator cuff injury.

June 2011 — minor update. The 2010/2011 QIPP options for local implementation have been added to this topic. 

July 2010 — minor update. In people at risk of cardiovascular adverse events, ibuprofen up to 1200 mg per day or naproxen up to 1000 mg per day are recommended as first-line NSAIDs.

August to November 2008 — this is a new CKS topic. The evidence-base has been reviewed in detail, and recommendations are clearly justified and transparently linked to the supporting evidence.

Update

New evidence

Evidence-based guidelines

No new evidence-based guidelines since 1 October 2022.

HTAs (Health Technology Assessments)

No new HTAs since 1 October 2022.

Economic appraisals

No new HTAs since 1 October 2022.

Systematic reviews and meta-analyses

No new systematic reviews published since 1 October 2022.

Primary evidence

No new primary evidence which reaches the CKS threshold for inclusion published since 1 October 2022.

New policies

No new national policies or guidelines since 1 October 2022.

New safety alerts

No new safety alerts since 1 October 2022.

Changes in product availability

No changes in product availability since 1 October 2022.

Goals and outcome measures

Goals

To support primary healthcare professionals to:

  • Make an accurate diagnosis of the underlying cause of the pain.
  • Make an accurate assessment (e.g. of severity, limitation of function).
  • Provide appropriate treatment in primary care settings.
  • Arrange referral to secondary care or other specialist services, if appropriate.
  • Provide appropriate self-management advice to patients.

Outcome measures

No outcome measures were found during the review of this topic.

Audit criteria

No audit criteria were found during the review of this topic.

QOF indicators

No QOF indicators were found during the review of this topic.

QIPP - Options for local implementation

 No QIPP indicators were found during the review of this topic.

NICE quality standards

No NICE quality standards were found during the review of this topic.

Background information

What causes shoulder pain?

Shoulder pain can be due to a number of causes. These include [Mitchell et al, 2005; Gray, 2016]:

  • Pain arising from the shoulder:
    • Rotator cuff disorders.
    • Frozen shoulder.
    • Instability disorders.
    • Acromioclavicular joint disorders.
    • Glenohumeral joint osteoarthritis.
    • Inflammatory arthritis.
    • Paget's disease.
    • Avascular necrosis of the humeral head.
    • Septic arthritis.
  • Pain arising from elsewhere:
    • Referred pain from the neck (for example, degenerative disc disease), diaphragm (for example, in intra-abdominal sepsis), heart (for example, ischaemic heart disease), or lungs (for example apical lung cancer).
    • Polymyalgia rheumatica.
    • Malignancy (for example, metastases to the humeral head). 
    • Early herpes zoster (C5-T1 dermatomes).

The four most common causes of shoulder pain in primary care are [Artus, 2014]:

  • Rotator cuff disorders.
  • Glenohumeral disorders (frozen shoulder or glenohumeral osteoarthritis).
  • Acromioclavicular joint disease.
  • Referred neck pain.

What are rotator cuff disorders?

  • The term 'rotator cuff' refers to the group of muscles and tendons that surround and stabilise the shoulder joint [BMJ Best Practice, 2018].
  • Rotator cuff disorders cause subacromial shoulder pain, which is felt in the top and lateral side of the shoulder [Kulkarni, 2015; Gray, 2016].
  • Rotator cuff disorders include rotator cuff tendinopathy and partial rotator cuff tears [Mitchell et al, 2005; Kulkarni, 2015; Gray, 2016].
    • Rotator cuff tendinopathy is also known as shoulder impingement. Impingement occurs between the acromion and the rotator cuff tendons (which can be intact or torn). As the diagnosis of other causes of subacromial pain (such as supraspinatus tendinopathy, tendinitis, and subacromial bursitis) is essentially the same as rotator cuff tendinopathy, these are referred to as rotator cuff tendinopathy in this CKS topic [Kulkarni, 2015].
    • Rotator cuff tears can occur through either trauma (such as shoulder dislocation) or attrition (such as chronic degeneration), and can be either symptomatic or asymptomatic [BMJ Best Practice, 2018].
  • Subacromial shoulder pain from rotator cuff disorders is the most common cause of shoulder pain presenting to primary care [Kulkarni, 2015; Gray, 2016].
    • It typically affects people between the ages of 35 and 75.
    • There may be a history of repetitive movements at or above shoulder height, or of heavy lifting.
    • It can occur in athletes, workers who perform overhead activities and the elderly [NZGG, 2004].
  • Diagnosis is based on the history and examination [Kulkarni, 2015].
    • There is pain on the top and lateral side of the shoulder, which is made worse by raising the arm (for example when lifting a full kettle) or with overhead activities. There may be pain at night. On examination, active movements are painful and may be restricted, whereas passive movements tend to be full but painful [Mitchell et al, 2005; BMJ Best Practice, 2018].
    • There may be [Mitchell et al, 2005; Kulkarni, 2015]:
      • A painful arc of movement between 70-120 degrees of abduction (however research has not shown this to be sensitive or specific as a clinical sign).
      • Pain on abduction with the thumb down, which is worse against resistance.
  • It is important to distinguish traumatic rotator cuff tear from other causes of subacromial shoulder pain.
    • Success of surgical repair is adversely affected by delayed diagnosis [Gray, 2016].
    • People with traumatic tears are usually active young to middle-aged adults and have usually sustained a dislocation or traction-type injury (where the shoulder is forcibly stretched).
    • On examination there is usually severe pain and significant weakness. There may be a positive 'drop arm test', where the person cannot support the weight of the affected arm when it is abducted to 90 degrees.
    • Urgent referral to orthopaedics is required [Kulkarni, 2015; Gray, 2016].

What is frozen shoulder?

  • Frozen shoulder is a painful condition that leads to stiffness and disability [Rangan, A., 2015]. It is characterised by a progressive restriction of both active and passive shoulder movement [BMJ Best Practice, 2021], and typically affects people aged between 40-60 years [Gray, 2016]. 
  • Frozen shoulder can be primary (idiopathic) or secondary [Rangan, A., 2015].
    • Secondary frozen shoulder is associated with trauma, rotator cuff disease, cardiovascular disease, hemiparesis, diabetes, and thyroid dysfunction [Rangan, A., 2015; BMJ Best Practice, 2021].
    • It has been estimated that the prevalence of frozen shoulder in people with diabetes is around 13 percent [Zreik, 2016].
    • Symptoms are often more severe and resistant to treatment compared to people without diabetes [Dias et al, 2005; Karbowiak, 2022].
    • Frozen shoulder is more common in women and in people who have previously had frozen shoulder in the opposite arm [Kelley, 2013]. 
  • Frozen shoulder progresses through three overlapping phases [Rangan, A., 2015; Karbowiak, 2022]:
    • Painful phase (lasts 2-9 months). There is progressive pain on movement. The pain can become severe and disturb sleep.
    • Stiffness phase (lasts 4-12 months). The pain becomes less severe but is present at the end of the range of movement. Stiffness remains and there is reduction in the range of shoulder movements. Function can be substantially limited. 
    • Resolution phase (lasts 12-42 months). Gradual improvement in range of movement with less stiffness.
  • As there is overlap between phases, frozen shoulder has been more recently classified into 'pain predominant' and 'stiffness predominant' phases [Hanchard, 2012; Rangan, A., 2015]. 
  • Although symptoms usually settle in 18-24 months, there are reports of residual pain and stiffness persisting for several years in some people [BMJ Best Practice, 2021].
  • The diagnosis of frozen shoulder is clinical [Kelley, 2013; Rangan, A., 2015; Rangan, 2016; BMJ Best Practice, 2021] and has the following features:
    • Gradual onset of pain in the deltoid region, with worsening shoulder stiffness, which may make activities of daily living difficult [Mitchell et al, 2005].
    • Restricted active and passive external rotation, and pain at the end of external rotation. There is a global limitation of active and passive range of shoulder movements with a capsular pattern (that is, a disproportionately severe reduction in passive external rotation).
    • Physical examination manoeuvres may be helpful to exclude other potential causes for a decreased range of motion in the shoulder (for example, rotator cuff disorders or cervical spine injury).
    • No obvious crepitus on movement.
    • X-rays are not routinely necessary. If performed, the glenohumeral joint should appear normal [Rangan, 2016]. 
    • It is important to note that restricted passive external rotation is seen in other disorders, such as glenohumeral osteoarthritis, avascular necrosis, and dislocation.

What are instability disorders?

  • Shoulder instability is the abnormal movement of the head of the humerus. It can cause pain, subluxation (partial loss of contact between joint surfaces) or dislocation (the complete loss of contact between joint surfaces which needs a reduction manoeuvre) [Thangarajah, 2015]. 
  • Dislocation can be acute, recurrent, or persistent (locked).
  • Shoulder joint laxity is the asymptomatic movement of the humeral head on the glenoid cavity at the upper end of the normal range of movement.
  • Laxity can occur without instability, and vice versa, or the two may co-exist [NZGG, 2004; Thangarajah, 2015].
  • Shoulder instability has been classified as follows [Thangarajah, 2015; Gray, 2016; Brownson, 2015; Noorani, 2019]:
    • Traumatic structural instability. This is when the shoulder is dislocated by an external force. As the shoulder does not heal in the correct anatomical position, or structures do not heal properly, the person is more susceptible to recurrent dislocations and further damage. 96% of shoulder dislocations are due to trauma. 
    • Atraumatic structural instability. This is common in adolescent females with hypermobile joints. If the shoulder is lax, it becomes unstable spontaneously or following a minor injury. 
    • Muscle patterning. This is due to unbalanced muscle recruitment around the shoulder, and there is no structural abnormality. The person may have poor posture. 
    • These patterns can co-exist, or develop one after another with time. 
  • Instability disorders usually occur in younger people, typically those aged less than 25 to 35 years of age [Brownson, 2015; Oxford Shoulder Clinic, 2010; Noorani, 2019].
  • The person may give a history of the shoulder feeling that it moves partly or completely 'out of joint', and may be concerned their shoulder may dislocate during certain activities or sports. There may be other non-specific symptoms such as shoulder ache or intermittent clicking. If the instability is longstanding, there may be hand or arm weakness, tingling, or numbness from proximal nerve traction [Brun, 2012a]. Few tests of shoulder instability have been shown to be useful in practice [Thangarajah, 2015]. Often, there are no abnormalities on examination [Brun, 2012a]. The Beighton's score can be used if hypermobility is suspected [Thangarajah, 2015].
  • Anterior dislocation is the most common traumatic dislocation and is most often seen in people aged 15-30. There is another peak in women aged over 80. Typically the person presents with significant pain following a fall onto an outstretched hand. On examination, there is loss of the normal contour of the shoulder [NZGG, 2004]. 
  • A younger age at first dislocation is associated with a higher risk of repeated dislocation. A person aged less than 20 has a 90% chance of having a further dislocation [NZGG, 2004; Brownson, 2015]. Men are more likely than women to have a recurrence, and people with hyperlaxity are more likely to suffer dislocation than those without hyperlaxity [Olds, 2015].
  • Injuries that may be associated with traumatic shoulder dislocations include nerve injury (for example, to the axillary nerve), greater tuberosity fracture and rotator cuff tear [Brownson, 2015; Noorani, 2019]. These injuries are more commonly seen in older people [Paxton, 2014].
  • Posterior shoulder dislocation makes up 2-4% of all shoulder dislocations. The diagnosis is often missed at first presentation. This may be because the health care professional or the person with shoulder pain did not feel the injury was sufficient to dislocate the shoulder; because examination findings are more subtle than seen in anterior dislocation; or because only one X-ray view of the shoulder has been performed. Early diagnosis is important as there is often an associated fracture of the humeral head, which, if undetected, can lead to osteoarthritis or avascular necrosis [Jacobs, 2015]. Posterior dislocation can be caused by indirect trauma, with the arm flexed at the shoulder in adduction and internal rotation. It is important to consider posterior dislocation in someone with shoulder pain following a seizure. On examination, the person may hold the shoulder in adduction and internal rotation. There may be an abnormal shoulder contour. 
  • Osteoarthritis is a complication of shoulder instability [Thangarajah, 2015]. 

What are acromioclavicular joint disorders?

These include acromioclavicular osteoarthritis and acromioclavicular joint injuries. 

  • Acromioclavicular osteoarthritis is more common than osteoarthritis of the glenohumeral joint, but is often asymptomatic.
    • It usually presents in people over 60 [Gray, 2016]. The person may have previously sustained an acromioclavicular joint sprain.
    • Weightlifting (putting the joint under repetitive strain) is a risk factor [Brun, 2012a].
    • Symptoms include tenderness over the acromioclavicular joint, which is made worse by raising the arm up high or when the arm is brought across the body [Gray, 2016]. It may also be associated with rotator cuff disorders [Mitchell et al, 2005].
  • Acromioclavicular joint injuries can involve stretching or tearing of the acromioclavicular or coracoclavicular ligaments and subluxation or dislocation of the acromioclavicular joint. They most commonly occur in men aged 20-50 years, and often result from a fall onto the point of the shoulder during sporting activity (for example, rugby and skiing) [NZGG, 2004].
  • Acromioclavicular joint injuries are classified as [NZGG, 2004]:
    • Grade I: Intact joint with minor tear of the acromioclavicular ligaments.
    • Grade II: Up to 50% vertical subluxation of the clavicle with rupture of the acromioclavicular ligament and stretching of the coracoclavicular ligaments. 
    • Grade III: more than 50% vertical subluxation of the clavicle with complete rupture of both acromioclavicular and coracoclavicular ligaments. 
  • On examination there may be tenderness localised to the AC joint, limited range of movement due to pain, high arc pain, or a positive cross arm test  [NZGG, 2004; Oxford Shoulder Clinic, 2010; Brun, 2012a].
  • There may be asymmetry of the shoulder contours following AC joint injury, however, this may only become apparent weeks after an injury [NZGG, 2004]. 

What is glenohumeral joint osteoarthritis?

  • Glenohumeral joint osteoarthritis is a less common cause of chronic shoulder pain, but can cause significant pain and disability [Thomas, M., 2016]. As the shoulder is a non-weight-bearing joint, primary arthritis is rare. Secondary causes account for the majority of cases, such as following trauma, chronic rotator cuff tear, and systemic arthritis [Goud, 2008]. 
  • Glenohumeral joint osteoarthritis is more common in people older than 60 [Mitchell et al, 2005].
  • It presents with deep joint pain and restriction of activities, such as putting on a jacket [Mitchell et al, 2005; Gray, 2016]. Other joints may be involved [Thomas, M., 2016].
  • On examination there is a global restriction in active and passive shoulder movement, with a severe loss of passive external rotation [Mitchell et al, 2005; Thomas, M., 2016].
  • X-rays of the shoulder (AP and axillary views) are needed to confirm the diagnosis [Thomas, M., 2016]. 

How common is shoulder pain?

  • Shoulder pain is common [Artus, 2017]. Estimates indicate that shoulder pain is the third most common musculoskeletal presentation in primary care, with approximately 1% of adults presenting with new shoulder pain annually [Artus, 2014].
  • Lifetime prevalence estimates vary widely but have indicated that up to 67% of people may develop shoulder pain [Hodgets, 2021].

What is the prognosis?

  • Shoulder pain can result in significant morbidity and functional disability [Eubank, 2021], and has been associated with poor general health [Hodgets, 2021].
  • While some patients present with minor and short-lived symptoms, chronicity and recurrence can be common [Eubank, 2021; Artus, 2014].
  • Treatment success is likely dependent upon an accurate assessment of the cause of the shoulder pain [Kauta, 2004]. However, the overall prognosis of shoulder pain is highly variable and dependent on the cause [van der Wint, 2019; Hodgets, 2021].
    • Approximately 40 to 50% of people with shoulder pain experience persistent pain 6 to 12 months after consulting their primary care clinician.
    • Long-term prognosis can be influenced by age, educational level, symptom duration, number of previous episodes of shoulder pain, disability scores, impacts upon range of motion, and the presence of multisite pain.

Diagnosis

How should I assess someone with a painful shoulder?

  • Assessment of a person with shoulder pain involves taking a history and performing an examination to help identify the underlying cause, and any red flags for a serious underlying condition.
  • History should include: 
    • Characteristics of the pain: onset; duration; site of maximal pain; whether the pain is felt at rest, on movement, or both; night pain and whether pain affects sleeping position; radiation; and aggravating and relieving factors.
    • Neck pain or other upper limb pain.
    • Stiffness.
    • Instability: any past history of dislocations or concerns that the shoulder might come out of joint during certain movements. If there is a past history of dislocation, ask if the first dislocation was caused by an injury; how old the person was when the shoulder was first dislocated; number of dislocations; and the direction in which the shoulder dislocated. 
    • Trauma: any preceding injury or strenuous activity. Interval between this and the onset of pain. Mechanism of injury: site injured, arm position at the time of injury, and any subjective experiences, such as feeling something 'pop out'. 
    • Functional impairment: whether the dominant or non-dominant arm is affected; any restriction of activities; or effects on work or sport.
    • Pain in other joints. 
    • Systemic features such as fever, night sweats, weight loss, rash, and respiratory symptoms.
    • Neurological symptoms.
    • Occupation and sporting activities, for example, involving repetitive arm movements and long periods of elevation.
    • Past history of shoulder or musculoskeletal problems.
    • Past medical history, in particular significant comorbidities such as diabetes, stroke, ischaemic heart disease, malignancy (such as lung or breast cancer), and gastrointestinal or renal disease.
    • Current medications. Consider potential adverse drug reactions, such as statin myopathy.
    • Family history.
  • Examination should include a comparison of both shoulders: 
    • As an initial test, ask the person to place the palms of their hands at the base of the neck with elbows pointing laterally, and then to put their arms down and try to put the back of the hands between the shoulder blades. This also involves a mobility assessment of other joints including elbow and wrist. 
    • Inspect the shoulder from the front, side, and behind for muscle wasting, swelling, deformity, or bruising.
    • Palpate the shoulder bones (clavicle, proximal humerus, and scapula) and joints (sternoclavicular, acromioclavicular, glenohumeral), looking for tenderness, warmth, swelling, and crepitus.
    • Assess active, passive, and resisted movement of the shoulder joint. Assess flexion, extension, abduction, adduction, and internal and external rotation.
    • Look for the painful arc of abduction, if this is possible within the limits of discomfort (pain between 70–120 degrees of active abduction), and, if present, check if there is pain on abduction with the thumb down and if the pain is worse against resistance. 
    • Perform the cross-arm test. This is positive if there is pain over the acromioclavicular joint when the person raises the affected arm to 90 degrees, then actively adducts the shoulder.
    • Examine the neck, arms, axillae, and chest wall for possible sources of referred pain. Assess the range of movement of the cervical spine. 
    • Perform a neurological examination if indicated. 
    • If neck movement reproduces the pain, and the shoulder pain is thought to be referred from the neck, see the CKS guidelines on Neck pain - acute torticollis Neck pain - cervical radiculopathy Neck pain - non-specific Neck pain - whiplash injury for further assessment and management.
    • If shoulder movement reproduces the pain, and the pain is thought to arise from the shoulder, consider following the Oxford University Hospitals algorithm for the diagnosis of shoulder problems in primary care www.ouh.nhs.uk. 
  • Perform investigations, if appropriate.
    • Investigations should be guided by the suspected cause. (Blood tests and radiography are not usually indicated as part of a primary care assessment of shoulder pain - see Initial management for details).
  • Perform blood tests if malignancy, polymyalgia rheumatica, or inflammatory arthritis are suspected.
  • Consider testing for diabetes if this is suspected in a person with shoulder pain (frozen shoulder is more common in people with diabetes).
  • Consider anteroposterior and lateral shoulders X-rays if:
    • There is a history of trauma.
    • The person is not improving with conservative treatment, or symptoms are lasting more than four weeks. 
    • Movement is significantly restricted.
    • There is severe pain.
    • Any red flags are present.  
    • Arthritis is suspected.
  • Ultrasound and MRI scans are not usually requested in primary care.

Basis for recommendation

General approach to management

  • CKS advises a general approach to assessment and has not included a large number of specific tests (e.g. for individual muscles, tendons, or joints) because:
    • There is no evidence that any specific test is both valid and reliable for the diagnosis of shoulder injuries [NZGG, 2004].
    • A Cochrane review concluded there was no sufficient evidence for which tests should be used to identify shoulder impingement in primary care, and that tests to diagnose acromioclavicular joint disease are of limited value. There is no evidence that a reduced range of passive external rotation is accurate in diagnosing frozen shoulder [Hanchard, 2011].  
    • A review concluded that no single test in the shoulder examination can be reliably recommended to make a diagnosis [Hegedus, 2012].
    • There is a lack of consensus on diagnostic criteria and clinical assessment, even between musculoskeletal specialists. Additionally, mixed shoulder disorders are common and over-differentiation between the many diagnostic categories is unlikely to alter standard primary care management [Mitchell et al, 2005].

Investigations in primary care

  • CKS does not recommend routine investigations in primary care:
    • Review articles [Mitchell et al, 2005; Artus, 2014] advise that malignancy and systemic illnesses are infrequent causes of shoulder pain, and that blood tests, such as full blood count and erythrocyte sedimentation rate should be performed if there are red flag symptoms or examination findings. 
    • Due to the high prevalence of frozen shoulder in people with diabetes, and the high prevalence of diabetes in people with frozen shoulder, the authors of a meta-analysis advised that diabetes should be considered in people presenting with frozen shoulder [Zreik, 2016]. 
  • The recommendations regarding when to consider performing X-rays are based on a review article [Artus, 2014] and the BESS/BOA guideline [Kulkarni, 2015].
    • The diagnosis of frozen shoulder is essentially clinical. X-rays are not routinely needed unless to exclude arthritis [NZGG, 2004; Rangan, 2016]. The BESS/BOA guideline recommends performing an X-ray to confirm the diagnosis of glenohumeral osteoarthritis and exclude other diagnoses such as avascular necrosis [Rangan, A., 2015]. 
    • The BESS/BOA guideline for the management of subacromial shoulder pain advises considering an X-ray if the person is not improving with conservative treatment [Kulkarni, 2015]. 
    • The BESS/BOA guideline recommends that diagnostic ultrasound is performed in secondary care, rather than in primary care [Kulkarni, 2015]. A normal ultrasound of the rotator cuff does not exclude serious causes of shoulder pain, such as malignancy. The accuracy of ultrasound in diagnosis is dependent on the clinician performing the examination. Partial tears are a common report finding; however, these can be asymptomatic or a false positive examination finding. Therefore if a tear is found on ultrasound, it does not always correlate with the person's symptoms or mean surgery will definitely be needed. The BESS/BOA guideline recommends ultrasound findings should be interpreted by orthopaedic surgeons and considered along with the person's symptoms and previous management. A review also recommends that ultrasound and MRI scans are usually secondary care investigations [Artus, 2014]. 

Management

Management

From age 16 years onwards.

How should I manage a person with shoulder pain initially?

  • Refer a person urgently to secondary care if any red flags are identified in the history or examination. These include: 
    • Trauma, pain and weakness, or sudden loss of ability to actively raise the arm (with or without trauma): suspect acute rotator cuff tear.
    • Any shoulder mass or swelling: suspect malignancy.
    • Red skin, painful joint, fever, or the person is systemically unwell: suspect septic arthritis.
    • Trauma leading to loss of rotation and abnormal shape: possible shoulder dislocation.
    • New symptoms of inflammation in several joints: suspect inflammatory arthritis. 
  • Also consider urgent investigations and/or referral to secondary care if there: 
    • Are systemic symptoms such as fever, night sweats, weight loss, or new respiratory symptoms. 
    • Is undiagnosed severe shoulder pain or severe restriction of movement.  
    • Is a history of trauma and the person is being seen acutely.
  • Refer a person urgently to secondary care for same-day emergency assessment for: 
    • Suspected joint infection (red skin, person has fever or is systemically unwell).
    • Unreduced dislocation (trauma, epileptic fit, or electric shock leading to abnormal shoulder shape and loss of rotation).
    • Acute trauma, depending on clinical judgement.
  • Refer a person urgently for suspected:
    • Malignancy (past history of cancer, symptoms or signs of cancer, mass or swelling, unexplained deformity, lymphadenopathy). Follow the local 2-week referral pathway.
    • Acute rotator cuff tear caused by trauma (trauma, pain, and weakness). Refer urgently to orthopaedics. The person should be seen in the next available specialist shoulder clinic. 
    • Inflammatory arthritis. Follow the local rheumatology referral pathway.
    • Neurological lesion (unexplained wasting, significant motor or sensory deficit). Discuss with neurology, neurosurgery or orthopaedics, depending on clinical judgement. 
  • For suspected intrinsic shoulder disorders, follow the Oxford University Hospitals guidelines for the diagnosis, treatment and referral of common shoulder problems in primary care www.ouh.nhs.uk [Oxford Shoulder Clinic, 2010].
    • If there is a history of instability (the shoulder has ever come out of joint, or the person is concerned their shoulder may dislocate during certain activities or sport) a diagnosis of instability is likely.
    • If there is pain and tenderness over the AC joint, high arc pain, or a positive cross-arm test, acromioclavicular joint disease is likely.
    • If there is reduced passive external rotation, a diagnosis of a glenohumeral joint disorder (frozen shoulder or arthritis) is likely. 
    • If there is a painful arc of abduction or pain on abduction with the thumb down, a diagnosis of a rotator cuff disorder is likely.  Note that a young person with a history of trauma and a loss of abduction needs urgent referral to exclude acute cuff tear. 
  • For the initial management of a person with shoulder pain without any red flags:
    • Explain the diagnosis to the person and offer an information leaflet, such as the Versus Arthritis Shoulder Pain leaflet.
    • Prescribe appropriate analgesia. Offer paracetamol first line. If paracetamol is ineffective, consider an oral NSAID (for example, ibuprofen) or codeine. Consider whether there are any contraindications and which drug has a more favourable balance of benefits and risks for the person. If there is no early benefit from an oral NSAID, discontinue its use. For prescribing information on NSAIDs, information on minimizing associated risks and when to consider gastroprotection, see the CKS topic on NSAIDs - prescribing issues. For prescribing information on codeine, see the CKS topic on Analgesia - mild-to-moderate pain.
    • Consider the person's work and leisure activities. 
    • Advise the person to carry out normal activities as much as they are able to. Consider a short time off work, for example, one week, if there appears to be a direct link between this and the shoulder pain. Advise rest from activities that worsen the shoulder pain for a few weeks, such as sport, and, if relevant, advise 'light duties' at work for a few weeks. 
    • Explain that although common shoulder problems tend to be self-limiting, the rehabilitation period can be at least 6 months.
  • Consider reviewing the person after 2 weeks of initial management:
  • Refer early to secondary care if:
    • There is recurrent shoulder instability.
    • There is severe post-traumatic pain.
    • The pain is having a significant impact on the person, for example, on work or if they are an athlete.
  • Refer to secondary care if pain and function are not improving following conservative treatment for 3 months.
  • Consider referral to a specialised musculoskeletal clinic, for example, for provision of physiotherapy or corticosteroid injection. However, if orthopaedic referral is indicated, do not delay this referral.

Basis for recommendation

Red flags
  • The recommendations for features in the history and examination which require urgent investigations, and/or referral are based on expert opinion in UK guidelines [Royal College of Surgeons, 2014; Kulkarni, 2015; Brownson, 2015; Rangan, A., 2015; Thomas, M., 2016], the Oxford University Hospitals guideline for the diagnosis of shoulder problems in primary care [Oxford Shoulder Clinic, 2010], a New Zealand guideline [NZGG, 2004], and in review articles [Mitchell et al, 2005; Artus, 2014].
  • The recommendation to refer a person urgently to secondary care for same day assessment following acute trauma, depending on clinical judgement, is based on what CKS considers best clinical practice.
  • The recommendation to discuss a suspected neurological lesion with neurology, neurosurgery or orthopaedics, depending on clinical judgement, is also based on what CKS considers to be best clinical practice. 
Diagnosis of intrinsic shoulder disorders
Initial management of a person with shoulder pain without any red flags, and referral to secondary care
  • This is based on two review articles on the management of shoulder pain in primary care [Artus, 2014; Gray, 2016].
    • One of these reviews advises on the importance of pain control in allowing the person to return to normal activities as early as possible [Artus, 2014].
    • This review also recommends that restarting normal activities and returning to work early in the course of shoulder pain may help reduce the chance of restricted function in the long-term [Artus, 2014].
    • A 2015 survey of shoulder pain management in primary care in the UK identified that pain medications are routinely prescribed for shoulder pain [Artus, 2017].
  • There is evidence that physiotherapy and home exercises may provide some improvement for people with shoulder pain in the short term [Buchbinder et al, 2003; Green et al, 2003].
  • The optimum timing of referral for physiotherapy is not clear [Artus, 2014].
  • There is no definite evidence on how soon people with shoulder pain should be reviewed. The recommendations for timing of review are based on suggestions from a review article [Artus, 2014].
  • The recommendations on general condition management are closely aligned with the core values and behaviours described in the NHS England musculoskeletal core capabilities framework for first point of contact practitioners [NHS Health Education England, 2018].
Intermediate care
  • Guidance around referral to intermediate care aims to reduce referrals to hospital while ensuring that people are directed towards the most appropriate services and clinicians working at the interface between primary and secondary care [NICE, 2017].
  • Intermediate care services provide rapid assessment, diagnosis and treatment of people with a variety of musculoskeletal problems. Healthcare professionals involved may include physiotherapists, GPs with a special interest in musculoskeletal problems, and clinical nurse practitioners [Royal College of Surgeons, 2014].

Scenario: Frozen shoulder

From age 16 years onwards.

How should I manage a person with frozen shoulder?

  • Explain the diagnosis and what the person should expect. 
    • Pain is the main problem initially. It can be worse in bed and disturb sleep. 
    • The pain gradually improves, but stiffness slowly worsens and becomes the main problem. The stiffness then gradually resolves.
    • Frozen shoulder is usually self-limiting, but it can take months to years to resolve. 
  • Advise on activity modification and pain control. 
    • When the shoulder is painful, continue to use the arm to maintain movement and ease spasm. Avoid movements which worsen the pain. This may require time off work or away from leisure activities.
    • Take analgesia as advised.
    • Hot packs may be helpful. 
    • In bed, support the arm with pillows (to prevent rolling onto the affected shoulder).
  • Consider the following in the person's management:
    • Analgesia.
    • Supervised physiotherapy/domestic exercise programme.
    • Intra-articular corticosteroid injection.
    • Referral to secondary care.
  • Consider following a step-up approach, starting with non-invasive treatments, and moving on to invasive ones, if required.
    • Consider the severity of symptoms and the impact it is having on the person (how much this is restricting work or leisure activities) when planning treatment.  
  • Offer analgesia, particularly in the early, painful phase.
    • Consider paracetamol first line. Regular dosing is more effective than 'as required use'. For further information see the CKS topic Analgesia - mild-to-moderate pain. 
    • If paracetamol is ineffective, consider an oral NSAID (for example, ibuprofen) or codeine. Consider any contraindications and which drug has a more favourable balance of benefits and risks for the person. 
    • If there is no early benefit from an oral NSAID, discontinue its use. 
    • For prescribing information on NSAIDs, information on minimizing associated risks, and when to consider gastroprotection, see the CKS topic on NSAIDs - prescribing issues.
    • For prescribing information on codeine, see the CKS topic on Analgesia - mild-to-moderate pain.
  • Consider referral for physiotherapy. Start physiotherapy as early as possible in the course of frozen shoulder. Ensure adequate analgesia is provided.
    • The course of physiotherapy is usually 6 weeks unless the person cannot tolerate the exercises or the physiotherapist advises earlier referral to secondary care.
    • Physiotherapy can be continued for a further 6 weeks if there is improvement in the person's symptoms or function.
    • Physiotherapy can involve education and advice, exercises, manual therapy, thermotherapy, and electrotherapy. A corticosteroid injection or acupuncture may be performed by the physiotherapist.  
  • Consider an intra-articular (glenohumeral) corticosteroid injection early in the course of frozen shoulder if there is no, or slow, progress with conservative treatment. 
    • This can be done in primary care if the expertise is available, otherwise refer to orthopaedics, physiotherapy, or intermediate care. 
    • Intra-articular corticosteroid injections should be administered by an appropriately trained and skilled person.
    • Discuss potential adverse effects when obtaining informed consent. These include infection (rare if sterile technique is used), tendon rupture, 'post-injection flare of pain', and hyperglycaemia in people with diabetes.
    • Although a number of steroid preparations are available, triamcinolone or methylprednisolone are the preferred option for many specialists.
    • Local anaesthetic (e.g. lidocaine) is frequently used in addition to corticosteroid. Advise the person to rest the injected joint as much as is practical for 24 hours following the injection. Advise them that if there is severe pain after the injection, especially if it is associated with fever or shivering, to seek urgent medical advice.
    • Vigilant self-monitoring of blood sugars is recommended for people with diabetes following steroid injection, as transient hyperglycaemia may occur for 24–48 hours.
    • Where feasible, steroid injections should be deferred until diabetes is well-controlled and the risk of infection should be carefully considered.
    • If a corticosteroid injection does not produce the expected benefit, the diagnosis should be reviewed or referral considered.
  • Do not give a corticosteroid injection if:
    • The person has previously had an intra-articular corticosteroid injection from an experienced healthcare practitioner, with minimal or no benefit.
    • The pain has settled and stiffness is the predominant feature.
    • Corticosteroid injection is contraindicated (e.g. infection, sensitivity to local anaesthetic, or adjacent osteomyelitis).
  • Refer to secondary care:
    • If the person has had pain and or stiffness for up to 3 months and has not benefited from conservative measures. Consider referral earlier than this if symptoms are severe, or there is considerable impact on the person's quality of life.
    • If there is uncertainty about the diagnosis.
    • For consideration of a corticosteroid injection, if this cannot be done in primary care.

Basis for recommendation

Approach to management
  • The explanation and advice for people with frozen shoulder is based on expert opinion in review articles [Rangan, 2016; BMJ Best Practice, 2021]. 
  • Frozen shoulder is usually managed in primary care where the management options include analgesia, physiotherapy, and intra-articular corticosteroid injections. A small proportion of people with frozen shoulder are referred to orthopaedics.
    • Management options in secondary care include manipulation of the shoulder joint under general anaesthetic and capsular release [Rangan, 2016]. 
  • The management options are based on the BESS/BOA guideline for the management of frozen shoulder [Rangan, A., 2015].
  • There is a lack of evidence for the efficacy of conservative and surgical treatments for frozen shoulder [Rangan, A., 2015; Rangan, 2016].
  • A step-up approach in terms of treatment invasiveness should be considered. Management should depend on the degree of symptoms and functional impairment, and involve shared-decision making [BMJ Best Practice, 2021; Rangan, 2016]. 
Analgesia
  • BESS/BOA recommend analgesia for the pain of frozen shoulder.
    • Pain relief is particularly important in the initial painful stage [Karbowiak, 2022] and may be useful in helping compliance with physical therapy [Moskowitz and Blaine, 2005].
    • NSAIDs have been shown to reduce pain in people with acute capsulitis [Rangan, A., 2015]. Expert opinion in a review article [BMJ Best Practice, 2021] states that prescribing NSAIDs early in the disease course may also reduce the disease severity. The effectiveness of paracetamol and codeine for the management of frozen shoulder is not known [Rangan, A., 2015]. However, CKS advises that the choice of analgesia should be based on clinical judgement, taking into account the severity of the person's symptoms and comorbidities. 
      • For more information on the adverse effect profile and contraindications to prescribing NSAIDs, see the CKS topic on NSAIDs - prescribing issues.
    • CKS has not recommended the use of stronger opioids in primary care as long-term treatment should be undertaken with caution because of the potential for dependence [Iannotti and Kwon, 2005; Moskowitz and Blaine, 2005].
Physiotherapy
  • Physiotherapy has been shown to have some benefit in the management of frozen shoulder [Rangan, 2016]. This benefit is seen in the short-to-medium term [Rangan, A., 2015]. 
  • The recommendation to start physiotherapy as early as possible is based on expert opinion in a review article [BMJ Best Practice, 2021].
  • The length of physiotherapy treatment is as recommended by BESS/BOA [Rangan, A., 2015] and in an expert opinion review article [Karbowiak, 2022].
  • It is unclear which mode of physiotherapy is best for people with frozen shoulder as there is insufficient evidence [Rangan, A., 2015]. For people who have already had an intra-articular corticosteroid injection or physiotherapy, short-wave diathermy plus stretching and high grade mobilization techniques may be beneficial. 
  • Physiotherapists provide advice and education, which people with frozen shoulder find very important [Jones, 2013]. 
  • A supervised exercise programme has been found to give a faster improvement in range of movement than a simple home programme of gentle exercise [NZGG, 2004].
    • CKS recommends ensuring adequate analgesia is provided so that the person is able to tolerate the movements involved. 
Intra-articular corticosteroid injection
  • Intra-articular steroid injection has been shown to have some benefit in the management of frozen shoulder [Rangan, A., 2015; Rangan, 2016; BMJ Best Practice, 2021; Karbowiak, 2022].
    • A single injection has been shown to have a short term improvement in pain, function and range of movement compared with placebo [Maund, 2012], and possibly compared with manual therapy [Page, 2014], in people who have had frozen shoulder for up to one year [Rangan, 2016].
    • Intra-articular steroid injection has been associated with better pain relief compared with subacromial injections [Chen, 2019].
    • Physiotherapy following steroid injection may be more effective than physiotherapy on its own [Rangan, 2016; Karbowiak, 2022].
  • The BESS/BOA guideline recommends intra-articular corticosteroid injection as a treatment option for frozen shoulder [Rangan, A., 2015].
    • This guideline recommends that clinicians follow a step-up approach in terms of treatment invasiveness, and also take into account the severity of symptoms and impact these are having on the person.
  • Expert opinion in a review article recommends considering a steroid injection for people who still have pain despite starting physiotherapy and a trial of NSAIDS (if there are no contraindications) [BMJ Best Practice, 2021]. It also recommends that a steroid injection could be offered before starting physiotherapy so as to allow early intensive physiotherapy.
  • Expert opinion in another review article [Rangan, 2016] suggests that physiotherapy in the pain-predominant phase is likely to include an intra-articular steroid injection, and a New Zealand guideline also recommends considering an intra-articular steroid injection in the painful phase of frozen shoulder [NZGG, 2004].
  • Current evidence does not allow definite conclusions to be made as to the clinical situations for which steroid injection (with or without physiotherapy) is most likely to be an effective treatment [Rangan, A., 2015]. The BESS/BOA guideline advises to refer to orthopaedics if primary care management is ineffective [Rangan, A., 2015]. It does not comment on whether or not injection should be repeated if it has been beneficial. However, there is some evidence that repeated injections can cause tendon damage [Dean, 2014].
  • Corticosteroid injection is not recommended in the stiffness predominant phase, as the inflammatory stage is over [Dias et al, 2005]. 
  • Intra-articular steroid injection may be technically difficult [NZGG, 2004]; therefore, it should only be performed in primary care if the expertise is available.
    • A 2015 survey of shoulder pain management in primary care in the UK identified that GPs with a special interest in musculoskeletal medicine, and those with specific musculoskeletal/shoulder training, were more likely to provide corticosteroid injections [Artus, 2017].
  • Choice of corticosteroid and use of lidocaine:
    • CKS found no guidelines informing the choice of corticosteroid.
    • A number of corticosteroid preparations are listed in the British National Formulary [BNF, 2022], but feedback from expert reviewers suggested that triamcinolone or methylprednisolone are the most commonly used.
    • Expert opinion in a review article [BMJ Best Practice, 2021] suggests using triamcinolone acetonide.
    • Feedback from expert reviewers also suggested that local anaesthetic is commonly used in conjunction with a corticosteroid. Reasons given for this included that local anaesthetic can give immediate pain relief, can help confirm the diagnosis, and may decrease the risk of steroid atrophy by diluting the corticosteroid. Expert opinion in a review article [BMJ Best Practice, 2021] suggests using triamcinolone acetonide administered with a local anaesthetic.
  • Adverse effects: 
    • CKS advises discussion of adverse effects when considering corticosteroid injections. These include infection (rare if sterile technique used), tendon rupture, 'post-injection flare of pain', and hyperglycaemia in people with diabetes [NZGG, 2004; DTB, 1995; ABPI, 2022].
    • Advice for the person to rest for 24 hours following the injection, and seek medical advice if there is severe pain, especially if associated with fever, is recommended by the Drug and Therapeutics Bulletin review [DTB, 1995]. 
    • A New Zealand guideline advises monitoring blood glucose in people with diabetes after corticosteroid injection as this may be elevated for 24–48 hours [NZGG, 2004]. The risk may be highest in the first 24 hours after treatment [Shin, 2020].
  • Feedback from expert reviewers suggests that it may be preferable to defer steroid injections until diabetes is well controlled, and that the risk of infection should be carefully considered in a person with diabetes. 
Referral
  • The recommendations for when to refer someone with frozen shoulder to orthopaedics are based on the BESS/BOA guideline for the diagnosis and management of frozen shoulder [Rangan, A., 2015], and the Oxford University Hospitals guideline for the diagnosis of shoulder problems in primary care [Oxford Shoulder Clinic, 2010].
Treatments not recommended
  • Oral corticosteroids are not recommended by BESS/BOA for the management of frozen shoulder in primary care [Rangan, A., 2015]. There is evidence they may reduce pain early in the disease, but only in the short term [BMJ Best Practice, 2021]. Intra-articular steroid injection has been shown to be more effective [BMJ Best Practice, 2021]. Therefore this CKS guideline does not recommend the prescription of oral corticosteroids for frozen shoulder. 

Scenario: Rotator cuff disorders

From age 16 years onwards.

How should I manage a person with a rotator cuff disorder?

  • Consider the following in the person's management:
    • Rest (in the acute phase).
    • Exercise/ physiotherapy.
    • Corticosteroid injection.
    • Referral to intermediate or secondary care.
  • Adopt a shared-decision making approach when making management plans, and consider making treatment goals. 
  • Advise rest in the acute phase. Following this, activity should gradually be increased. 
  • Advise modification of activities that exacerbate symptoms (such as reaching overhead). Usual activities, within the limits of pain, should be restarted as soon as possible.
  • Offer analgesia.
    • Consider paracetamol first line. Regular dosing is more effective than 'as required use'. For further information see the CKS topic on Analgesia - mild-to-moderate pain
    • If paracetamol is ineffective, consider an oral NSAID (for example, ibuprofen) or codeine. Consider any contraindications and which drug has a more favourable balance of benefits and risks for the person. 
    • If there is no early benefit from an oral NSAID, discontinue its use. 
    • For prescribing information on NSAIDs, information on minimising associated risks and when to consider gastroprotection, see the CKS topic on NSAIDs - prescribing issues.
    • For prescribing information on codeine, see the CKS topic on Analgesia.
  • Refer for physiotherapy.
    • The course of physiotherapy is usually 6 weeks, unless the person cannot tolerate the exercises or the physiotherapist advises earlier referral to secondary care.
    • Physiotherapy may include postural correction, motor control retraining, stretching and strengthening of the rotator cuff and scapular muscles, and manual therapy.
    • Low intensity and high-frequency exercise, that is within the person's pain threshold, should be performed.
    • Physiotherapy can be continued for a further 6 weeks if there is improvement in the person's symptoms or function.
  • Consider a subacromial corticosteroid injection as part of shared decision making, and in conjunction with other primary care treatments, such as physiotherapy.
    • This can be done in primary care if the expertise is available, otherwise refer to orthopaedics, physiotherapy, or intermediate care. 
    • Intra-articular corticosteroid injections should be administered by an appropriately trained and skilled person.
    • Discuss potential adverse effects when obtaining informed consent. These include infection (rare if sterile technique used), tendon rupture, 'post-injection flare of pain', and hyperglycaemia in people with diabetes
    • Although a number of steroid preparations are available, triamcinolone or methylprednisolone are the preferred option for many specialists.
    • Local anaesthetic (e.g. lidocaine) is frequently used in addition to corticosteroid.
    • Advise the person to rest the injected joint as much as is practical for 24 hours following the injection. Advise the person that if there is severe pain after the injection, especially if associated with fever or shivering, to seek urgent medical advice. 
    • Vigilant self-monitoring of blood sugars is recommended for people with diabetes following steroid injection, as transient hyperglycaemia may occur for 24–48 hours.
    • Where feasible, steroid injections should be deferred until diabetes is well-controlled and the risk of infection should be carefully considered.
    • If a corticosteroid injection does not produce the expected benefit, review the diagnosis or consider referral.
  • Do not give more than two corticosteroid injections. 
    • Normally only one injection should be given (due to the risk of tendon damage from repeated injections).
    • A second injection may be given after 6 weeks, but only to people who have seen improvement following the first injection, and who need the pain relief to allow them to undergo physiotherapy. 
  • Do not give a corticosteroid injection if:
    • The person has previously received a corticosteroid injection from an experienced healthcare practitioner, with minimal or no benefit.
    • The person has previously had two injections in the same shoulder.
    • Corticosteroid injection is contraindicated (e.g. infection, sensitivity to local anaesthetic, adjacent osteomyelitis).
  • Consider referral to intermediate care, if appropriate and locally available:
    • Integrated Clinical Assessment and Treatment Services (ICATS) can offer: assessment, treatment programmes (non-surgical), and referral to secondary care.  
  • Refer to secondary care if:
    • The person has not benefited from at least 6 weeks of non-surgical treatment (for example has persistent pain and reduced function).
    • The diagnosis is uncertain.
  • Consider factors that could delay surgical treatment if referring to secondary care. For example, investigate and treat diabetes.
  • Refer urgently to secondary care:
    • If any red flag features are present, see Initial management. 
    • If acute rotator cuff tear caused by trauma is suspected (trauma, pain, and weakness). Ideally, the person should be seen in the next available specialist shoulder clinic. 

Basis for recommendation

Approach to management
  • There is a lack of well-designed trials evaluating the common treatments for shoulder disorders. Evidence for the efficacy of conservative and surgical treatments is limited. Therefore, the BESS/BOA guideline recommends shared decision-making in the person's management, and that the clinician considers the severity of symptoms and the impact that these are having on the person [Kulkarni, 2015].
  • Management options for subacromial shoulder pain are based on this guideline and a Royal College of Surgeons commissioning guide [Royal College of Surgeons, 2014].
  • Shared decision making and goal setting are also recommended by an Australian guideline for the management of rotator cuff syndrome in the workplace [UNSW, 2013].
  • Non-surgical treatments for rotator cuff disorders are often used together, rather than alone [UNSW, 2013; Diercks, R., 2014; Kulkarni, 2015].
Rest
  • A Dutch guideline recommends rest in the acute phase, and then to gradually increase activity [Diercks, R., 2014].
  • An Australian guideline for the management of rotator cuff syndrome in the workplace advises workers should restart their usual activities as soon as possible, within the limits of pain. This is important for avoiding future disability and improving function [UNSW, 2013]. 
Advice
  • New Zealand, Dutch, and UK guidelines [NZGG, 2004; Diercks, R., 2014; Kulkarni, 2015] and an article on the management of shoulder pain in primary care [Mitchell et al, 2005] recommend self-help advice and a discussion of physical contributory factors and analgesia.
  • A Dutch guideline recommends activity modifications for occupations that involve repetitive movements of the arm, prolonged use of the arms, and hand held vibration tools [Diercks, R., 2014].
    • Reduction of precipitating movements (particularly overhead activities) is advised, as this helps the person avoid the painful arc between 60 and 120 degrees [Burbank et al, 2008]. 
Analgesia
  • Oral NSAIDs may reduce pain in the short term for people with acute tendonitis or subacromial bursitis [Kulkarni, 2015].
    • A Dutch guideline recommends using NSAIDs only if necessary during the acute phase, and for a maximum of 2 weeks [Diercks, R., 2014].
    • A New Zealand guideline recommends caution with NSAIDS in view of side effects, and recommends that paracetamol may give adequate pain relief [NZGG, 2004].
    • A systematic review found there was low to moderate grade evidence that NSAIDS reduced pain in people with rotator cuff tendinopathy, and moderate evidence that NSAIDS were as effective as a steroid injection for pain in the short term [Boudreault, J., 2014]. It also found that there is insufficient evidence regarding the use of NSAIDS in the moderate to long term.  
  • It is not known whether topical NSAIDS, paracetamol, or opioids  are beneficial for shoulder pain [Kulkarni, 2015].
    • As the evidence for the efficacy of conservative treatments for subacromial shoulder pain as a result of rotator cuff disorders, and the benefits of one over another, is limited, CKS advises that the choice of analgesia should be based on clinical judgement, taking into account the severity of the person's symptoms and comorbidities. 
  • For more information on the adverse effect profile and contraindications to prescribing NSAIDs, see the CKS topic on NSAIDs - prescribing issues.
  • CKS has not recommended the use of stronger opioids in primary care as long-term treatment should be undertaken with caution due to the potential for dependence [Iannotti and Kwon, 2005; Moskowitz and Blaine, 2005].
Physiotherapy/exercise
  • Australian, Dutch and UK guidelines recommend physiotherapy for subacromial shoulder pain [UNSW, 2013; Diercks, R., 2014; Kulkarni, 2015].
    • Evidence suggests physiotherapy may be beneficial for pain and improving function in people with mixed shoulder disorders.
    • The types of treatment are as recommended by a UK guideline [Royal College of Surgeons, 2014].
    • The level and intensity of exercises are as recommended by a Dutch guideline [Diercks, R., 2014].
  • A 2016 Cochrane review of manual therapy and exercise for rotator cuff disease found that only one of the 60 relevant trials included manual therapy and exercise which was consistent with current practice [Page, 2016]. 
    • This trial showed manual therapy and exercise improved function slightly more than placebo did at 22 weeks, and that there was little difference in pain.
    • The trial also provided low quality evidence that there was little difference in outcome when manual therapy and exercise were compared with corticosteroid injection. 
Corticosteroid injections
  • UK guidelines recommend that subacromial corticosteroid injection can be used in the management of subacromial shoulder pain [Kulkarni, 2015; Royal College of Surgeons, 2014].
    • They report that due to limited evidence, it is not known whether this is effective.
    • The guidelines state that corticosteroid injections can be considered along with rest, exercise, physiotherapy, and analgesia in the person's management, and that management should involve shared decision-making.
  • A Dutch guideline recommends a corticosteroid injection can be used in the first 8 weeks for severe pain, along with other conservative management options, such as physiotherapy [Diercks, R., 2014].
  • An Australian guideline advises subacromial corticosteroid injection for injured workers with persistent pain, or those making limited progress with an active non-surgical treatment programme [UNSW, 2013].
  • One meta-analysis found corticosteroid injection was no more effective than placebo injection at reducing pain at the 3 month follow up, and gave transient pain relief in a small number of people with rotator cuff tendinosis [Mohamadi A., 2016]. 
  •  An expert opinion review article provides treatment guidance recommending that steroid injections be used in conjunction with analgesia and mobilizing physiotherapy [BMJ Best Practice, 2018].
  • A review suggests if a person has seen little improvement following initial activity modification, analgesia, and physiotherapy, or if the person has significantly reduced function due to pain at the start, a corticosteroid injection may provide pain relief to allow a greater range of shoulder movement and tolerance of physiotherapy [Burbank et al, 2008]. 
Choice of corticosteroid and use of lidocaine
  • CKS found no guidelines informing the choice of corticosteroid. A number of corticosteroid preparations are listed in the British National Formulary [BNF, 2022], but feedback from expert reviewers suggested that triamcinolone or methylprednisolone are the most commonly used. 
  • Feedback from expert reviewers also suggested that local anaesthetic is commonly used in conjunction with a corticosteroid.
    • Reasons given for this included that local anaesthetic can give immediate pain relief, can help confirm the diagnosis, and may decrease the risk of steroid atrophy by diluting the corticosteroid.
    • An Australian guideline also advises the use of local anaesthetic for pain relief and to confirm that the needle had been placed accurately [UNSW, 2013]. 
Number of injections
  • The BESS/BOA and Royal College of Surgeons guidelines advise not to give more than two corticosteroid injections, and that normally only one injection should be given [Kulkarni, 2015; Royal College of Surgeons, 2014].  Evidence suggests that repeated frequent injections could cause damage to tendons [Dean, 2014].
Adverse effects
  • CKS advises discussion of adverse effects when considering subacromial corticosteroid injections. These include infection (rare if sterile technique used), tendon rupture, infections, 'post-injection flare of pain', skin atrophy, and hyperglycaemia in people with diabetes [DTB, 1995; NZGG, 2004; UNSW, 2013]. Feedback from an expert reviewer suggests discussion of other adverse effects including local tissue atrophy, flushing, and menstrual disorders in women. 
  • A New Zealand guideline advises monitoring blood glucose in people with diabetes after corticosteroid injection [UNSW, 2013], as this may be elevated for 24–48 hours [NZGG, 2004].
  • Feedback from expert reviewers suggests that it may be preferable to defer steroid injections until diabetes is well controlled, and that the risk of infection should be carefully considered in a person with diabetes. 
  • Advice for the person to rest for 24 hours following the injection, and seek medical advice if there is severe pain, especially if associated with fever, is recommended by a Drug and Therapeutics Bulletin review [DTB, 1995]. 
Intermediate care
  • Aims to reduce referrals to hospital while ensuring that people are directed towards the most appropriate services and clinicians working at the interface between primary and secondary care [NICE, 2017].
  • They provide rapid assessment, diagnosis, and treatment of people with a variety of musculoskeletal problems.
  • Healthcare professionals involved may include physiotherapists, GPs with a special interest in musculoskeletal problems, and clinical nurse practitioners [Royal College of Surgeons, 2014].
Referral to secondary care
Urgent referral to orthopaedics

Scenario: Instability disorders

From age 16 years onwards.

How should I manage a person with an instability disorder?

  • Refer immediately to an Emergency Department if acute dislocation is suspected. The person, or any bystanders, should not try to reduce the dislocation. There may be associated injuries; therefore, clinical assessment and radiological investigations should be performed, and reduction performed in a controlled environment. 
  • Following initial assessment and reduction of an acute traumatic shoulder dislocation: 
    • Encourage early mobilisation (as soon as the pain allows). 
    • Refer to physiotherapy. The course of physiotherapy is usually 4-12 weeks. 
    • Note that most people with an acute dislocation will have follow up in secondary care.
      • People aged under 25 should be seen by a shoulder surgeon within 6 weeks to consider whether surgical repair is appropriate.
      • People aged 25-40 should be assessed for symptoms of shoulder instability at 3-6 months, and further imaging and surgery considered if there are still symptoms.
      • People aged over 40 should have early imaging following a dislocation to check the integrity of the rotator cuff. If a rotator cuff tear is found, the person should be reviewed by a shoulder surgeon regarding the need for surgery.
  • Refer urgently to orthopaedics if a rotator cuff tear is suspected in a person who has recently sustained an acute shoulder dislocation. Suspect a rotator cuff tear if there is still pain and weakness 2-3 weeks following dislocation.
  • Refer to a shoulder surgeon:
    • Following first-time dislocation.
    • If the person has had recurrent painful dislocations.
    • If symptoms are impacting on the person's job or leisure activities. 
    • If the person has had a shoulder dislocation associated with an epileptic seizure.
    • If symptoms following shoulder dislocation are not improving with physiotherapy.
    • If there are associated injuries, such as fracture.
  • Refer to physiotherapy if:
    • Atraumatic shoulder instability is suspected. Refer to orthopaedics if physiotherapy is not beneficial. 
    • Muscle patterning instability is suspected. Refer to orthopaedics if the shoulder is still unstable despite physiotherapy.
  • Refer urgently if any red flag features are present.

Basis for recommendation

Acute dislocation
  • The recommendations regarding the management of a person with an acute shoulder dislocation are based on a BESS/BOA guideline [Brownson, 2015], which advises:  
    • The person should be referred immediately to an Emergency department for clinical assessment, radiological investigations, and reduction. X-rays (AP and lateral) are required to confirm the diagnosis of dislocation, the direction of the dislocation, and to check for greater tuberosity fracture. Thorough neurological and circulatory examinations should be performed before and after the shoulder is relocated. Adequate analgesia (with or without sedation) should be given. Relocation should only be performed by a clinician with adequate expertise. X-rays should be taken to check the shoulder has been correctly relocated, and whether any greater tuberosity fracture is in a satisfactory position.  
    • Further investigations, such as ultrasound or magnetic resonance imaging with arthrography (MRA), may be performed, depending on the age of the person; how likely it is they will need surgery; the number of previous dislocations; and whether there are signs of associated injuries.
    • People aged over 40 should have further imaging (such as ultrasound or MRA) to check rotator cuff integrity. This age group has an increased risk of rotator cuff tears, and early surgical repair may be required. 
  • The recommendations regarding the management of a person following reduction of a shoulder dislocation are based on a BESS/BOA guideline [Brownson, 2015].
    • Follow up should usually be through the fracture clinic in secondary care. 
    • Immobilisation for more than a week does not reduce the likelihood of having another dislocation. The person should mobilise the shoulder as soon as the pain allows. Early referral to physiotherapy may help with this. Physiotherapy aims to maintain the relocated shoulder joint so that the structures can heal properly and the joint and surrounding muscles can provide stability to the shoulder [Thangarajah, 2015]. 
    • The decision to perform surgery following first time dislocation is based on the risk of further dislocations (for example, age and sex); the likelihood of complications (such as post-traumatic osteoarthritis); occupation; and sporting activities. 
    • If a rotator cuff tear is suspected following dislocation, the person should be referred urgently to a shoulder surgeon. Expert opinion in a systematic review advises rotator cuff tear should be suspected in people who have persistent pain and weakness 2-3 weeks after dislocation [Gombera, M.M., 2014]. 
  • The BESS/BOA guideline proposes that following uncomplicated primary anterior dislocation [Brownson, 2015]:
    • People aged under 25 should be seen by a shoulder specialist before 6 weeks, and further imaging and arthroscopic anatomic repair considered. Young men have an especially high risk of recurrent dislocation. The evidence suggests arthroscopic repair is beneficial, especially if the person plays contact sports or has an occupation that involves overhead activities.  
    • People aged 25-40 should be reviewed at 3-6 months.  If there are symptoms, further imaging and repair should be considered. 
Referral
  • The recommendation to refer people with recurrent traumatic instability to orthopaedics is based on a BESS/BOA guideline [Brownson, 2015]. It recommends that surgery will be needed in most cases. It also recommends referral to a shoulder surgeon if physiotherapy is not successful following dislocation. A New Zealand guideline also recommends people with recurrent dislocation should be referred to an orthopaedic specialist to evaluate the need for surgical stabilisation [NZGG, 2004]. 
  • The recommendations to refer people with symptoms that are impacting on their job or leisure activities, and people with shoulder dislocation and seizures, to a shoulder surgeon is based on expert opinion in a review article [Thangarajah, 2015].
    • Dislocation and instability can occur following a seizure. Management can be difficult because the shoulder structures can be damaged after the person has had a surgical repair. The person should be managed by a multidisciplinary team, which should also include a neurologist and physiotherapist. 
  • The recommendation to refer people with suspected muscle patterning instability to physiotherapy is based on expert opinion in a review article [Thangarajah, 2015].
  • The recommendation to refer people with suspected atraumatic instability to physiotherapy, and to orthopaedics if this is not considered likely to be beneficial, is based on a BESS/BOA guideline [Noorani, 2019] and the Oxford University hospitals guidelines for the diagnosis of common shoulder problems in primary care [Oxford Shoulder Clinic, 2010]. 

Scenario: Acromioclavicular joint disorders

From age 16 years onwards.

How should I manage a person with an acromioclavicular joint disorder?

  • For osteoarthritis of the acromioclavicular joint:
    • Advise the person on activity modification, in particular, to avoid cross-body shoulder adduction (for example, playing golf), and to avoid heavy lifting.
    • Offer analgesia.
    • Consider referral to physiotherapy.
    • Consider corticosteroid injection if pain is severe.
    • Arrange an X-ray (if this has not already been done) and refer to orthopaedics if there is no, or only temporary, improvement following the above measures.
    • See the CKS topic on Osteoarthritis for further information regarding the management of osteoarthritis.
  • Analgesia
    • Consider paracetamol first line. Regular dosing is more effective than 'as required' use. For further information, see paracetamol prescribing issues in the CKS topic Analgesia - mild-to-moderate pain.
    • If paracetamol is ineffective, consider an oral NSAID (for example, ibuprofen) or codeine. Consider any contraindications and which drug has a more favourable balance of benefits and risks for the person.
    • If there is no early benefit from an oral NSAID, discontinue its use.
    • For prescribing information on NSAIDs, information on minimising associated risks and when to consider gastroprotection, see the CKS topic on NSAIDs - prescribing issues.
    • For prescribing information on codeine, see the CKS topic on Analgesia.
  • For acromioclavicular joint injuries
    • Refer for an X-ray or to accident and emergency, depending on clinical judgement, if there is a history of trauma see Initial management.
    • If the acromioclavicular joint pain is caused by a mild sprain following an acute injury (See grade I or II sprains):
      • Initially, advise rest, consider providing a sling for 5–7 days, and offer analgesia (see analgesia section above).
      • Start gentle mobilisation and strengthening when the person is more comfortable. Consider referral to physiotherapy.
      • Advise the person to resume normal activities as tolerated, but avoid heavy lifting and contact sports for 8–12 weeks.
      • Refer to orthopaedics if symptoms are not settling following conservative management. Consider earlier referral for certain groups for whom shoulder pain is particularly disabling (for example, athletes involved in overhead sports or people involved in heavy manual labour).
    • For more severe injuries (grade 3 or higher dislocations), or if there is diagnostic uncertainty, discuss with, or refer to, orthopaedics.

Basis for recommendation

Management of acromioclavicular joint osteoarthritis
  • For further information on the management of osteoarthritis, see the CKS topic on Osteoarthritis.
  • The recommendations regarding management options for acromioclavicular osteoarthritis are based on the Oxford guideline for treatment and referral of shoulder pain [Oxford Shoulder Clinic, 2010] and a review of the management of chronic shoulder pain [Burbank et al, 2008].
    • The review recommends steroid injection as an option for the short-term relief of severe pain.
Management of acromioclavicular joint injuries
  • The recommendation to consider referral for X-ray or to accident and emergency, if there is a history of trauma, is based on a review of the management of shoulder problems in primary care [Mitchell et al, 2005] and a New Zealand guideline on the diagnosis and management of soft tissue shoulder injuries [NZGG, 2004]. It is also based on the BESS/BOA guidelines which advise on the management of red flags for shoulder pain [Thomas, M., 2016] and the Oxford guideline for treatment and referral of shoulder pain [Oxford Shoulder Clinic, 2010].
    • X-ray allows assessment of the extent of damage, which then determines the management [NZGG, 2004].
    • A New Zealand guideline recommends people with mild AC joint injury (grade I or II sprains), can be managed with a sling (for up to 1 week) and analgesia. Following this, the person can gradually return to normal activities. Contact sports and heavy lifting should not be performed for at least 8-12 weeks.
    • A review and best practice guideline recommend a rehabilitation programme following the acute phase which includes strengthening exercises [Reid, D., 2012], and a review of the management of shoulder injuries in general practice also advises referral to physiotherapy [Brun, 2012a].
    • A review of the management of shoulder pain in primary care advises that, unless there is a significant traumatic dislocation, acromioclavicular joint problems usually resolve with rest and simple analgesia [Mitchell et al, 2005].
    • The New Zealand guideline advises considering referral to an orthopaedic specialist if a sprain is suspected and there is an inadequate improvement after 3 months, or earlier for certain groups who may benefit from surgical stabilisation [NZGG, 2004].
    • Feedback from an expert reviewer suggests that rheumatologists may manage this group of people with corticosteroid injections.
    • A review of shoulder injuries in general practice also advises consideration of a steroid injection if conservative measures are ineffective [Brun, 2012a]. 
  • People with grade III acromioclavicular joint sprains (Grade III dislocations) can be managed conservatively [NZGG, 2004; Reid, D., 2012].
    • The New Zealand guideline recommends if conservative measures are not successful, to refer to orthopaedics at 3 months. Some groups, such as athletes, may require surgical stabilisation. It advises that more severe injuries usually require surgical management.
  • CKS recommends, as part of best clinical practice, that for more severe injuries (grade III or higher dislocations), or if there is diagnostic uncertainty, to discuss with, or refer to, orthopaedics.

Scenario: Glenohumeral joint osteoarthritis

From age 16 years onwards.

How should I manage a person with glenohumeral joint osteoarthritis?

  • For information on the management of osteoarthritis, see the CKS topic on Osteoarthritis.
  • Management decisions should be based on the severity of symptoms and the impact on the person's function, quality of life, mood, occupation, and leisure activities.
  • Consider the following core treatments for everyone with osteoarthritis:
    • Education and advice Information and advice.
    • Strengthening exercises. Consider referral to physiotherapy.
    • Weight loss if the person is overweight.
    • Assistive devices for people who have specific problems with activities of daily living. Consider seeking expert advice from an occupational therapist or Disability Equipment Assessment Centre.
  • Consider the following for a person with glenohumeral osteoarthritis:
    • Analgesia.
    • Physical therapy. Consider referral to physiotherapy.
    • Corticosteroid injection.
  • Offer analgesia.
    • Consider paracetamol first line. Regular dosing is more effective than 'as required' use. For further information, see paracetamol prescribing issues in the CKS topic on Osteoarthritis.
    • If paracetamol is ineffective, consider a topical NSAID. For further information, see the Topical NSAID prescribing information in the CKS topic on Osteoarthritis - Topical NSAIDs.
    • If topical medicines are ineffective or unsuitable, consider an oral NSAID (for example, ibuprofen) or codeine. Consider any contraindications and which drug has a more favourable balance of benefits and risks for the person.
    • If there is no early benefit from an oral NSAID, discontinue its use.
    • For prescribing information on NSAIDs, information on minimising associated risks, and when to consider gastroprotection, see the CKS topic on NSAIDs - prescribing issues.
    • For prescribing information on codeine, see the CKS topic on Analgesia - mild-to-moderate pain.
  • Consider a corticosteroid injection for the short-term management of an acute exacerbation of pain due to joint inflammation, or for temporary symptom control, for example, if surgery has been delayed. For further information, see Intra-articular corticosteroids prescribing information in the CKS topic on Osteoarthritis.
  • Refer to secondary care if:
    • Conservative measures do not control pain.
    • Stiffness is having a significant impact on the person.
    • The diagnosis is uncertain.
  • If a referral is needed for consideration of surgery, check that the person wishes to be referred.
  • Refer before there is prolonged and established functional limitation or severe pain.

Basis for recommendation

Approach to management
  • This reflects the NICE guideline for care and management of osteoarthritis in over 16 year olds [NICE, 2022]. For further information, see the CKS topic on Osteoarthritis. The BESS/BOA guideline for the management of glenohumeral osteoarthritis also recommends shared-decision making [Thomas, M., 2016].
Core treatments
  • This is based on the recommendations by NICE for the management of osteoarthritis [NICE, 2022].
  • The European League Against Rheumatism (EULAR) recommends assistive devices and adaptations at home or work [Fernandes et al, 2013].
Management options
  • The recommendations regarding management options for glenohumeral osteoarthritis are based on the BESS/BOA guideline [Thomas, M., 2016].
Analgesia
  • The BESS/BOA guideline recommends using paracetamol for pain [Thomas, M., 2016].
    • There is a good amount of evidence from RCTs on the efficacy of paracetamol in people with osteoarthritis, but trial data is mainly from people with osteoarthritis of the hip or knee.
    • Paracetamol reduces pain in the short term compared with placebo, but appears to be less effective than oral standard or coxib nonsteroidal anti-inflammatory drugs (NSAIDs), especially in people with moderate-to-severe pain.
    • Paracetamol can be used alone or in conjunction with other analgesics, although there is a limited quantity and quality of data on the efficacy of paracetamol used in combination with other analgesic drugs.
    • Paracetamol is less likely than NSAIDs to cause gastrointestinal (GI) adverse events.
    • NICE guidelines state that paracetamol should only be used infrequently for short-term pain relief where other treatments have been ineffective, are contraindicated, or not tolerated [NICE, 2022].
  • Expert opinion from a review article [Machado, 2021] and the SIGN guideline Management of chronic pain [SIGN, 2019] recommend considering the use of topical NSAIDs in people with chronic pain due to musculoskeletal conditions, particularly in those who cannot tolerate oral NSAIDs.
    • Topical NSAIDs do not increase the risk of gastrointestinal adverse events.
  • The BESS/BOA and NICE guidelines recommend that oral NSAIDs can be considered [Thomas, M., 2016; NICE, 2022].
    • A large amount of clinical trial evidence supports the efficacy of both standard NSAIDs and COX-2 selective NSAIDs in reducing the pain and stiffness of osteoarthritis, with the majority of studies reflecting short-term use compared with placebo and involving knee or hip joint osteoarthritis [Machado, 2021].
    • There is no strong evidence to suggest a consistent benefit over paracetamol, although some people may obtain greater symptom relief from NSAIDs.
    • No clinically important results about the effects of oral NSAIDs compared with opioids were found.
    • All NSAIDs, irrespective of COX-1 and COX-2 selectivity, are associated with significant morbidity and mortality due to adverse effects on the gastrointestinal, renal, and cardiovascular system [Machado, 2021].
  • The BESS/BOA guidelines recommend that codeine can be considered if NSAIDs are ineffective [Thomas, M., 2016]. NICE guidelines state that codeine can be considered if used infrequently for short-term pain relief where other treatments have been ineffective, are contraindicated, or not tolerated [NICE, 2022].
    • The evidence supporting the use of opioid analgesia in osteoarthritis is poor, but available trial data suggests that, compared with placebo, opioid analgesics reduce pain in people with osteoarthritis of the knee.
    • There is a lack of trials that compare the efficacy and safety of opioids with other symptomatic treatments for osteoarthritis, and there are virtually no good studies using opioids in people with peripheral joint osteoarthritis, so the benefits of opioids in different types of osteoarthritis remains unclear.
    • There is little evidence to suggest that increasing the opioid dose improves the effect.
    • There are also few data comparing different opioid formulations or routes of administration.
    • Adverse effects of opioids are a concern, especially in elderly people.
    • The BESS/BOA and NICE guidelines do not recommend codeine for long-term use due to adverse effects and risk of dependence [Thomas, M., 2016; NICE, 2022].
Physical treatments
  • The evidence for adjunctive physical (non-pharmacological) treatments is not strong. However, NICE recommended these modalities because they are safe and widely used, and those that can be part of the person's self-management plan are easily affordable by most people [NICE, 2022].
Corticosteroid injection
  • There is no evidence for the routine use of corticosteroid injections in the management of glenohumeral osteoarthritis [Gross, 2013].
  • The BESS/BOA guideline recommends that they may be useful in the short-term management of an acute exacerbation of pain due to joint inflammation, or for temporary symptom control, for example, if surgery is delayed [Thomas, M., 2016].
Treatments that are not recommended
  • The BESS/BOA guideline recommends acupuncture can be used in the management of glenohumeral osteoarthritis in primary care.
    • There is some evidence that acupuncture is an effective treatment for chronic shoulder pain [Lathia, 2009], however, the people studied had different shoulder conditions (for example frozen shoulder, rotator cuff tear).
    • There is some evidence that acupuncture is effective for osteoarthritis of the knee. For further information, see the CKS topic on Osteoarthritis Management.
    • NICE concluded that acupuncture has insufficient evidence of cost-effectiveness for any provision by the NHS [NICE, 2022].
Referral to secondary care
  • The criteria for referral is based on the BESS/BOA guideline [Thomas, M., 2016]. If symptoms do not improve with conservative treatments, then surgery may be considered. Options include arthroscopic interventions such as debridement, biological glenoid resurfacing with hemiarthroplasty, hemiarthroplasty, and total shoulder replacement.
    • People should not be excluded from referral based on age, sex/gender, smoking status, or other comorbidities [NICE, 2022].

Supporting evidence

This CKS topic is based on a number of clinical guidelines which are referenced throughout the topic. The rationale for the primary care diagnosis, management, and referral of shoulder pain is discussed in the relevant basis for recommendation sections. CKS has not summarized the evidence for secondary care investigations and management as they are outside the scope of this topic.

How this topic was developed

This section briefly describes the processes used in developing and updating this topic. Further details on the full process can be found in the About Us section and on the Clarity Informatics website.

Search strategy

A literature search was conducted for guidelines, systematic reviews and randomized controlled trials on primary care management of shoulder pain.

Search dates

April 2017 - October 2022

Key search terms

Various combinations of searches were carried out. The terms listed below are the core search terms that were used for Medline.

  • exp Shoulder Pain/ or shoulder pain.tw. or rotator cuff/ or rotator cuff.tw. or glenohumeral joint.tw. AND pain$.tw. exp Shoulder Impingement Syndrome/or adhesive capsulitis.tw. or exp Shoulder Joint/ or exp Joint Instability/ or (shoulder or frozen shoulder).tw. or exp Acromioclavicular Joint/

Sources of guidelines

Sources of systematic reviews and meta-analyses

  • The Cochrane Library:
    • Systematic reviews
    • Protocols
    • Database of Abstracts of Reviews of Effects
  • Medline (with systematic review filter)
  • EMBASE (with systematic review filter)

Sources of health technology assessments and economic appraisals

Sources of randomized controlled trials

  • The Cochrane Library:
    • Central Register of Controlled Trials
  • Medline (with randomized controlled trial filter)
  • EMBASE (with randomized controlled trial filter)

Sources of evidence based reviews and evidence summaries

Sources of national policy

Patient experiences

Sources of medicines information

The following sources are used by CKS pharmacists and are not necessarily searched by CKS information specialists for all topics. Some of these resources are not freely available and require subscriptions to access content.

Stakeholder engagement

Our policy

The external review process is an essential part of CKS topic development. Consultation with a wide range of stakeholders provides quality assurance of the topic in terms of:

  • Clinical accuracy.
  • Consistency with other providers of clinical knowledge for primary care.
  • Accuracy of implementation of national guidance (in particular NICE guidelines).
  • Usability.

Principles of the consultation process

  • The process is inclusive and any individual may participate.
  • To participate, an individual must declare whether they have any competing interests or not. If they do not declare whether or not they have competing interests, their comments will not be considered.
  • Comments received after the deadline will be considered, but they may not be acted upon before the clinical topic is issued onto the website.
  • Comments are accepted in any format that is convenient to the reviewer, although an electronic format is encouraged.
  • External reviewers are not paid for commenting on the draft topics.
  • Discussion with an individual or an organization about the CKS response to their comments is only undertaken in exceptional circumstances (at the discretion of the Clinical Editor or Editorial Steering Group).
  • All reviewers are thanked and offered a letter acknowledging their contribution for the purposes of appraisal/revalidation.
  • All reviewers are invited to be acknowledged on the website. All reviewers are given the opportunity to feedback about the external review process, enabling improvements to be made where appropriate.

Stakeholders

  • Key stakeholders identified by the CKS team are invited to comment on draft CKS topics. Individuals and organizations can also register an interest to feedback on a specific topic, or topics in a particular clinical area, through the Getting involved section of the Clarity Informatics website.
  • Stakeholders identified from the following groups are invited to review draft topics:
    • Experts in the topic area.
    • Professional organizations and societies (for example, Royal Colleges).
    • Patient organizations, Clarity has established close links with groups such as Age UK and the Alzheimer’s Society specifically for their input into new topic development, review of current topic content and advice on relevant areas of expert knowledge.
    • Guideline development groups where the topic is an implementation of a guideline.
    • The British National Formulary team.
    • The editorial team that develop MeReC Publications.
  • Reviewers are provided with clear instructions about what to review, what comments are particularly helpful, how to submit comments, and declaring interests.

Patient engagement

Clarity Informatics has enlisted the support and involvement of patients and lay persons at all stages in the process of creating the content which include:

  • Topic selection
  • Scoping of topic
  • Selection of clinical scenarios
  • First draft internal review
  • Second draft internal review
  • External review
  • Final draft and pre-publication

Our lay and patient involvement includes membership on the editorial steering group, contacting expert patient groups, organizations and individuals.

Evidence exclusion criteria

Our policy

Scoping a literature search, and reviewing the evidence for CKS is a methodical and systematic process that is carried out by the lead clinical author for each topic. Relevant evidence is gathered in order that the clinical author can make fully informed decisions and recommendations. It is important to note that some evidence may be excluded for a variety of reasons. These reasons may be applied across all CKS topics or may be specific to a given topic.

Studies identified during literature searches are reviewed to identify the most appropriate information to author a CKS topic, ensuring any recommendations are based on the best evidence. We use the principles of the GRADE and PICOT approaches to assess the quality of published research. We use the principles of AGREE II to assess the quality of published guidelines.

Standard exclusions for scoping literature:

  • Animal studies
  • Original research is not written in English

Possible exclusions for reviewed literature:

  • Sample size too small or study underpowered
  • Bias evident or promotional literature
  • Population not relevant
  • Intervention/treatment not relevant
  • Outcomes not relevant
  • Outcomes have no clear evidence of clinical effectiveness
  • Setting not relevant
  • Not relevant to UK
  • Incorrect study type
  • Review article
  • Duplicate reference

Organizational, behavioural and financial barriers

Our policy

The CKS literature searches take into consideration the following concepts, which are discussed at the initial scoping of the topic.

  • Feasibility
    • Studies are selected depending on whether the intervention under investigation is available in the NHS and can be practically and safely undertaken in primary care.
  • Organizational and Financial Impact Analysis
  • Studies are selected and evaluated on whether the intervention under investigations may have an impact on local clinical service provision or national impact on cost for the NHS. The principles of clinical budget impact analysis are adhered to, evaluated and recorded by the author. The following factors are considered when making this assessment and analysis.
    • Eligible population
    • Current interventions
    • Likely uptake of new intervention or recommendation
    • Cost of the current or new intervention mix
    • Impact on other costs
    • Condition-related costs
    • In-direct costs and service impacts
    • Time dependencies
  • Cost-effectiveness or cost-benefit analysis studies are identified where available. 

We also evaluate and include evidence from NICE accredited sources which provide economic evaluations of recommendations, such as NICE guidelines. When a recommended action may not be possible because of resource constraints, this is explicitly indicated to healthcare professionals by the wording of the CKS recommendation.

Declarations of interest

Our policy

Clarity Informatics requests that all those involved in the writing and reviewing of topics, and those involved in the external review process to declare any competing interests. Signed copies are securely held by Clarity Informatics and are available on request with the permission of the individual. A copy of the declaration of interest form which participants are asked to complete annually is also available on request. A brief outline of the declarations of interest policy is described here and full details of the policy is available on the Clarity Informatics website. Declarations of interests of the authors are not routinely published, however competing interests of all those involved in the topic update or development are listed below. Competing interests include:

  • Personal financial interests
  • Personal family interest
  • Personal non-financial interest
  • Non-personal financial gain or benefit

Although particular attention is given to interests that could result in financial gains or losses for the individual, competing interests may also arise from academic competition or for political, personal, religious, and reputational reasons. An individual is not obliged to seek out knowledge of work done for, or on behalf of, the healthcare industry within the departments for which they are responsible if they would not normally expect to be informed.

Who should declare competing interests?

Any individual (or organization) involved in developing, reviewing, or commenting on clinical content, particularly the recommendations should declare competing interests. This includes the authoring team members, expert advisers, external reviewers of draft topics, individuals providing feedback on published topics, and Editorial Steering Group members. Declarations of interest are completed annually for authoring team and editorial steering group members, and are completed at the start of the topic update and development process for external stakeholders.

Competing interests declared for this topic:

None.

References

  • ABPI (2021) SPC for Adcortyl Intra-Articular/Intradermal Injection 10mg/ml. Electronic Medicines Compendium. Datapharm Communications Ltd. https://www.medicines.org.uk [Free Full-text]
  • Artus, M., Holt, T. and Rees, J. (2014) The painful shoulder: an update on assessment, treatment, and referral. British Journal of General Practice. 64(626), e593-e595.
  • Artus, M., van der Windt, D.A., Afolabi, E.K., et al. (2017) Management of shoulder pain by UK general practitioners (GPs): a national survey. BMJ Open 7(6), e015711. [Abstract] [Free Full-text]
  • BMJ Best Practice (2018) Rotator cuff injury. BMJ Best Practice. https://bestpractice.bmj.com
  • BMJ Best Practice (2021) Adhesive capsulitis. BMJ Best Practice. https://bestpractice.bmj.com
  • BNF (2022) British National Formulary. National Institute for Health and Care Excellence (NICE). https://bnf.nice.org.uk
  • Boudreault, J., Desmeules, P.T. and Roy, J. et al (2014) The efficacy of oral non-steroidal anti-inflammatory drugs for rotator cuff tendinopathy: a systematic review and meta-analysis. Journal of Rehabilitation Medicine. 46(4), 294-306.
  • Brownson, P., Donaldson, O., Fox, M., et al. (2015) BESS/BOA patient care pathways: traumatic shoulder instability. Shoulder & elbow 7(3), 214-226. [Free Full-text]
  • Brun, S. (2012a) Shoulder injuries: management in general practice. Australian Family Physician 41(4), 188-194. [Abstract]
  • Brun, S. (2012b) Initial assessment of the injured shoulder. Australian Family Physician 41(4), 217-220. [Abstract]
  • Buchbinder, R., Green, S. and Youd, J.M. (2003) Corticosteroid injections for shoulder pain (Cochrane Review). The Cochrane Library. John Wiley & Sons, Ltd. http://www.thecochranelibrary.com [Free Full-text]
  • Burbank, K.M., Stevenson, J.H., Czarnecki, G.R. and Dorfman, J. (2008) Chronic shoulder pain: part II. Treatment. American Family Physician 77(4), 493-497. [Abstract]
  • Chen, R., Jiang, C. and Huang, G. (2019) Comparison of intra-articular and subacromial corticosteroid injection in frozen shoulder: A meta-analysis of randomized controlled trials. Int J Surg 68, 92-103. [Abstract] [Free Full-text]
  • Dean, B.J., Franklin, S.L, Murphy R.J, Javaid, M.K. and Carr, A.J. (2014) Glucocorticoids induce a specific ion-channel mediated toxicity in tendon. British Journal of Sports Medicine 48(22), 1620-1626. [Abstract]
  • Dias, R., Cutts, S. and Massoud, S. (2005) Frozen shoulder. British Medical Journal 331(7530), 1453-1456. [Abstract] [Free Full-text]
  • Diercks, R., Bron, C., Dorrestijin, O. Meskers, C., et al. (2014) Guideline for diagnosis and treatment of subacromial pain syndrome: a multidisciplinary review by the Dutch Orthopaedic Association. Acta Orthopaedica 85(3), 314-322. [Abstract]
  • DTB (1995) Articular and periarticular corticosteroid injections. Drug & Therapeutics Bulletin 33(9), 67-70. [Abstract]
  • Eubank, B.H.F., Lackey, S.W., Slomp, M., et al. (2021) Consensus for a primary care clinical decision-making tool for assessing, diagnosing, and managing shoulder pain in Alberta, Canada. BMC Fam Pract 22(201), 201. [Abstract] [Free Full-text]
  • Fernandes, L., Hagen, K.B., Bijlsma, J.W., et al. (2013) EULAR recommendations for the non-pharmacological core management of hip and knee osteoarthritis. Annals of the Rheumatic Diseases 72(7), 1125-1135. [Abstract]
  • Gombera, M.M. and Sekiya, J.K. (2014) Rotator cuff tear and glenohumeral instability: a systematic review. Clinical Orthopaedics and Related Research 472(8), 2448-2456. [Abstract]
  • Goud, A., Segal, D., Hedayati P.  Pan, J.J. and Weissman, B.N. (2008) Radiographic evaluation of the shoulder. European Journal of Radiology(68), 2-15. [Abstract]
  • Gray, M., Wallace, A. and Aldridge, S. (2016) Assessment of shoulder pain for non-specialists. British Medical Journal Dec 7(355), i5783. [Abstract]
  • Green, S., Buchbinder, R. and Hetrick, S. (2003) Physiotherapy interventions for shoulder pain (Cochrane Review). The Cochrane Library. John Wiley & Sons, Ltd. http://www.thecochranelibrary.com [Free Full-text]
  • Gross, C., Dhawan, A. and Harwood, D. et al (2013) Glenohumeral joint injections: a review . Sports Health. 5, 153-159. [Abstract]
  • Hanchard, N.C.A., Lenza, M., Hardoll, H.H.G. and Takwoingi, Y. (2013) Physical tests for shoulder impingements and local lesions of bursa, tendon or labrum that may accompany impingement (Cochrane Review/Cochrane Intervention Protocol). Issue 4. John Wiley & Sons, Ltd. http://www.cochranelibrary.com [Free Full-text]
  • Hanchard, N.C.A., Goodchild, L.M., Thompson, J. O'Brien, T., et al. (2012) Evidence-based clinical guidelines for the diagnosis, assessment and physiotherapy management of contracted (frozen) shoulder: quick reference summary. Physiotherapy 98(2), 117-120. [Abstract]
  • Hegedus, Goode, A.P., Cook, C.E. et al. (2012) Which physical examination tests provide clinicians with the most value when examining the shoulder? Update of a systematic review with meta-analysis of individual tests. British Journal of Sports Medicine 46(14), 964-978. [Abstract]
  • Hodgets, C. and Walker, B. (2021) Epidemiology, common diagnoses, treatments and prognosis of shoulder pain: A narrative review. International Journal of Osteopathic Medicine 42, 11-19. [Free Full-text]
  • Iannotti, J.P. and Kwon, Y.W. (2005) Management of persistent shoulder pain: a treatment algorithm. American Journal of Orthopedics 34(12 Suppl), 16-23. [Abstract]
  • Jacobs, R.C., Meredyth, N.A. and Michelson, J.D. (2015) Posterior shoulder dislocations. British Medical Journal 350(Jan 28). [Abstract]
  • Jones, S., Hanchard, N., Hamilton, S. and et al (2013) A qualitative study of patients’ perceptions and priorities when living with primary frozen shoulder. BMJ Open. 3(9), e003452. [Abstract]
  • Karbowiak, M., Holme, T., Mirza, M. and Siddiqui, N. (2022) Frozen Shoulder. BMJ 377, e068547. [Abstract]
  • Kauta, N., De Vries, E., Du Plessis, J.P., et al. (2004) Assessment and management of shoulder pain at primary care level. S Afr Fam Pract 63(1), e1-e4. [Abstract] [Free Full-text]
  • Kelley, M.J, Shaffer, M.A., Kuhn J.E. et al. (2013) Shoulder pain and mobility deficits: adhesive capsulitis. Journal of Orthopaedic & Sports Physical Therapy 43(5). [Abstract] [Free Full-text]
  • Kulkarni, R., Gibson, J., Brownson, P., et al. (2015) BESS/BOA patient care pathways: subacromial shoulder pain. Shoulder & elbow 7(2). [Free Full-text]
  • Lathia, A.T., Jung, S.M. and Chen, L.X. (2009) Efficacy of acupuncture as a treatment for chronic shoulder pain. Journal of Alternative and Complementary Medicine 15, 613-618. [Abstract]
  • Machado, G.C., Abdel-Shaheed, C., Underwood, M. and Day, R.O. (2021) Non-steroidal anti-inflammatory drugs (NSAIDs) for musculoskeletal pain. BMJ 372, n104. [Abstract]
  • Innes, J.A., Dover, A.R. and Fairhurst K. (Eds.) (2018) Macleod's Clinical Examination. 14th edn. Edinburgh: Elsevier.
  • Maund, E., Suekarran, S., Neilson, A., et al. (2012) Management of frozen shoulder: a systematic review and cost-effectiveness analysis. Health Technology Assessment 16(11), 1-264. [Abstract]
  • Mitchell, C., Adebajo, A., Hay, E. and Carr, A. (2005) Shoulder pain: diagnosis and management in primary care. British Medical Journal 331(7525), 1124-1128. [Free Full-text]
  • Mohamadi A., Chan J.J., Claessen F.M., et al. (2016) Corticosteroid injections give small and transient pain relief in rotator cuff tendinosis: a meta-analysis.. Clinical Orthopaedics and Related Research 475(1), 232-243. [Abstract]
  • Moskowitz, R.W. and Blaine, T.A. (2005) An overview of treatment options for persistent shoulder pain. American Journal of Orthopedics 34(12 Suppl), 10-15. [Abstract]
  • NHS Health Education England (2018) Musculoskeletal core capabilities framework for first point of contact practitioners. NHS Health Education England. http://www.hee.nhs.uk [Free Full-text]
  • NICE (2017) Intermediate care including reablement [NG74]. National Institute for Health and Care Excellence. https://www.nice.org.uk [Free Full-text]
  • NICE (2022) Osteoarthritis in over 16s: diagnosis and management. National Institute for Health and Care Excellence. https://www.nice.org.uk [Free Full-text]
  • Noorani, A., Goldring, M., Jaggi, A., et al. (2019) BESS/BOA patient care pathways: Atraumatic shoulder instability. Shoulder Elbow 11(1), 60-70. [Abstract] [Free Full-text]
  • NZGG (2004) The diagnosis and management of soft tissue shoulder injuries and related disorders. New Zealand Guidelines Group. http://www.acc.co.nz
  • Olds, M., Ellis, R., Donaldson, K. et al. (2015) Risk factors which predispose first-time traumatic anterior shoulder dislocations to recurrent instability in adults: a systematic review and meta-analysis. British Journal of Sports Medicine 49(14), 913-922. [Abstract]
  • Oxford Shoulder Clinic (2010) Diagnosis of shoulder problems in primary care: guidelines on treatment and referral. Nuffield Orthopaedic Centre. http://www.ouh.nhs.uk [Free Full-text]
  • Page M.J., Green, S., Kramer S. et al. (2014) Manual therapy and exercise for adhesive capsulitis (frozen shoulder)(Cochrane Review). Issue 8. John Wiley & Sons, Ltd. http://www.cochranelibrary.com [Free Full-text]
  • Page, M.J., Green, S., McBain, B. et al. (2016) Manual therapy and exercise for rotator cuff disease (Cochrane Review). Issue 6. John Wiley & Sons, Ltd. http://www.cochranelibrary.com [Free Full-text]
  • Paxton, E. S., Dodsonm C.C. and Lazarus, M.D. (2014) Shoulder instability in oler patients. The Orthopedic Clinics of North America 45(3), 377-385. [Abstract]
  • Rangan, A., Hanchard, N. and McDaid, C. (2016) What is the most effective treatment for frozen shoulder? British Medical Journal 354(Aug 23). [Abstract]
  • Rangan, A., Goodchild, L. and Gibson, J. et al (2015) BESS/BOA patient care pathways: frozen shoulder . Shoulder & elbow. 7(4), 299-307. [Abstract]
  • Reid, D., Polson, K. and Johnson, L. (2012) Acromioclavicular joint separations grades I-III: a review of the literature and development of best practice guidelines. 42(8), 681-696. [Abstract]
  • Royal College of Surgeons of England (2014) Commissioning guide: subacromial shoulder pain. Royal College of Surgeons. https://www.rcseng.ac.uk [Free Full-text]
  • Shin, W.Y., An, M.J., Im, N.G., et al. (2020) Changes in Blood Glucose Level After Steroid Injection for Musculoskeletal Pain in Patients With Diabetes. Annals of Rehabilitation Medicine 44(2), 117-124. [Abstract] [Free Full-text]
  • SIGN (2019) Management of chronic pain. Scottish Intercollegiate Guidelines Network. http://www.sign.ac.uk [Free Full-text]
  • Thangarajah, T. and Lambert, S. (2015) Management of the unstable shoulder. British Medical Journal. 350(May 28), 2537. [Abstract]
  • Thomas, M., Bidwi, A., Rangan, A. et al. (2016) BESS/BOA patient care pathways: glenohumeral osteoarthritis. Shoulder & elbow 8(3), 203-214. [Abstract]
  • UNSW (2013) Clinical Practice Guidelines for the Management of Rotator Cuff Syndrome in the Workplace. University of New South Wales. https://rcs.med.unsw.edu.au [Free Full-text]
  • van der Windt, D.A., Burke, D.L., Babatunde, O., et al. (2019) Predictors of the effects of treatment for shoulder pain: protocol of an individual participant data meta-analysis. Diagnostic and Prognostic Research 3(15). [Abstract] [Free Full-text]
  • Zreik, N.H., Malik, R.A., Charalambos P.C. et al. (2016) Adhesive capsulitis of the shoulder and diabetes: a meta-analysis of prevalence. Muscles, Ligaments and Tendons Journal 6(1), 26-34. [Abstract]
Change privacy settings