Diagnostic management of non-traumatic shoulder pain in adults and medical and surgical management of rotator cuff tendinopathy

Background

Shoulder pain is a common reason for consulting a general practitioner. The main cause of shoulder pain is rotator cuff disease, which accounts for around two out of every three cases. The prevalence of rotator cuff disease increases with age. It does not always cause symptoms, but can limit activities of daily living, lead to time off work and impair quality of life.

In a 2014 study, the French national health insurance system noted an increase in rotator cuff surgeries, with regional disparities. Furthermore, among patients who underwent rotator cuff surgery, 23% had not received appropriate medical treatment in the year leading up to their operation, and 36% had not had a preoperative MRI or CT arthrogram.

The HAS has issued two sets of guidelines to promote best practice:

• A first set of guidelines (link), published in 2023, which describes initial medical management, from diagnosis of the cause of the shoulder pain to the medical and functional treatment of rotator cuff disease, including the prevention of such disorders;
• A second set of guidelines (link), published in 2026, which set out the surgical indications for rotator cuff tendinopathy with a tear, the necessary preoperative imaging, prognostic factors, and pre- and postoperative rehabilitation.

These guidelines are supplemented by:

• Three toolkit guides:
  • Toolkit guide “Diagnostic management of non-traumatic shoulder pain in adults” (link), which outlines the main diagnoses to be considered in cases of acute and persistent shoulder pain, details (with diagrams) the clinical assessment to be performed for shoulder pain and, in particular, in cases of shoulder pain without stiffness;
  • Toolkit guide “Management of rotator cuff tendinopathies” (link), which defines the role of imaging and medical and functional treatments in rotator cuff disease;
  • Toolkit guide “Management of rotator cuff tendinopathy with a tear – the role of surgery” (link), which summarises the guidelines with the same title (in French only);

• Two report templates:
  • A template for a surgical report on a patient who has undergone surgery for rotator cuff tendinopathy with a tear (link);
  • Templates for preoperative imaging reports for rotator cuff disease, depending on the imaging technique (X-ray, ultrasound, MRI, CT arthrogram) (link).

Diagnostic management of non-traumatic shoulder pain in adults

In cases of non-traumatic shoulder pain, apart from rotator cuff disease, a number of other diagnoses may be considered: other mechanical conditions (osteoarthritis), infection or tumours, rheumatic conditions (inflammatory or microcrystalline arthritis) or referred pain.

The clinical assessment of the patient is essential for both diagnosis and treatment. This includes, in particular, specific questioning about the shoulder, visual inspection, palpation and assessment of joint range of motion (both active and passive) in both shoulders, a neurological examination of the upper limbs, an examination of the cervical spine and a combination of tests. Pain and function may be assessed using approved questionnaires in French (such as the Western Ontario Rotator Cuff Index).

The diagnostic hypotheses vary depending on how long the symptoms have been present and the clinical context. The diagnoses to be considered in the event of persistent shoulder pain are described in the algorithm below. It will also be necessary to eliminate differential diagnoses.

Medical management of rotator cuff tendinopathy

The first-line treatment for painful shoulders caused by degenerative lesions, with or without a tear, is medical and functional. Initial treatment includes:

• maintenance of tolerable activities;
• analgesics (paracetamol ± step 2) ± NSAIDs (unless contraindicated);
• education about the condition, advice and reassurance;
• management of loss of independence in older people.

If symptoms persist beyond 4 to 6 weeks, following X-rays of the painful shoulder:

• adjust pharmacological treatment;
• start physiotherapy;
• consider administering one or two corticosteroid injections into the shoulder (spaced at least 3 weeks apart, with details recorded in the report). Repeating injections is not recommended if there is no lasting improvement.

The decision to prescribe an ultrasound or MRI will be made on the basis of the clinical reassessment. Referral to a shoulder specialist (rheumatologist, physical medicine and rehabilitation specialist, sports medicine specialist or orthopaedic surgeon) is useful if symptoms do not improve.

Role of surgery in the management of rotator cuff tendinopathy

In rotator cuff tendinopathy without a tear, there is currently no evidence to support the use of surgery, and it should only be considered if a properly managed medical and functional treatment programme has failed. 

In cases of functional impairment or persistent, debilitating pain, surgery may be considered for degenerative rotator cuff tears after at least 6 months of medical and functional treatment. Consequently, not all tears, even full-thickness ones, require surgery.

A traumatic tear in an otherwise healthy tendon requires immediate surgical assessment, particularly in active patients. 

Preoperative imaging includes an X-ray of the shoulder (anteroposterior views in three rotations and a Lamy false-profile view) and cross-sectional imaging (MRI, CT arthrogram or MRI arthrogram).

Before determining whether surgery is indicated, the patient is assessed holistically and their biopsychosocial context is taken into account. Complex cases are treated using a multidisciplinary approach.

Where surgery is indicated, anatomical tendon repair is the gold-standard technique. It is most often performed arthroscopically. Associated procedures performed with the repair are determined on a case-by-case basis.

Where surgery is indicated, rehabilitation is essential both pre- and postoperatively. The rehabilitation and convalescence protocol depends on the specific procedure performed. It requires collaboration between the physiotherapist and the surgeon, including information sharing (particularly the surgical report and postoperative precautions).

Imaging is not routinely carried out postoperatively, unless there is an unfavourable outcome.

Current practice 

To provide context for its work on the “diagnostic management of non-traumatic shoulder pain in adults and the management of rotator cuff tendinopathy” and the “management of rotator cuff tendinopathy with a tear: the role of surgery”, the HAS conducted two studies using the French National Health Data System (SNDS). 

The first study notes that a large number of patients underwent isolated acromioplasty in the second half of 2022 (more than 3,600 adults aged 40 and over), even though, based on current evidence, this surgery has not been shown to be beneficial. Although the study focuses only on patients who underwent surgery, it suggests that the medical management of patients with non-traumatic tendinopathy without a tear is incomplete, as one in three patients who underwent surgery had not received physiotherapy in the 18 months prior to surgery, and only one in two patients had received at least one subacromial infiltration of cortisone derivatives. A reduction in the use of surgery and an improvement in medical and functional management are expected following publication of the guidelines. 

The second study found that, among the more than 26,000 adults who underwent rotator cuff repair surgery in France in the second half of 2022, many patients received inadequate preoperative care in the 18 months leading up to the operation: two in three patients did not receive the recommended medical and functional treatment (physiotherapy and infiltrations of cortisone derivatives), while one in three patients did not undergo the recommended imaging tests (standard X-rays accompanied by an MRI scan, a CT arthrogram or an MRI arthrogram). With regard to postoperative management in the 6 months following surgery, while imaging tests were carried out too frequently and too early (one in two patients underwent standard X-rays within a median of 40 days), the rate of postoperative physiotherapy was satisfactory (94% of patients).

Key figures from these studies can be found in the infographics in French. Detailed reports on these studies are included in the supplementary documents.

The SNDS data processing and analysis programmes are available on this page.