There's no cure, but there are plenty of treatment options to try. At least frozen shoulder doesn't last forever.

Frozen shoulder is one condition I dread diagnosing. Not only will your patient have to suffer a painful, debilitating condition for at least 12-18 months but it’s also a pain (literally!) to treat. The DTB have recently published a review on frozen shoulder (DTB 2015;53:90). Despite no new magic cures, it provides a useful summary of management strategies.

What is frozen shoulder?

Unfortunately there is no standard definition; whilst the term adhesive capsulitis is sometimes used interchangeably with frozen shoulder there is actually no evidence of adhesions within the shoulder capsule.

  • Estimated prevalence is 8-10% working age people, most commonly age 40-60y.
  • Commoner in diabetics, where it lasts longer and is often treatment resistant.
  • More commonly seen after upper limb injury.
  • Aetiology is unknown but appears to involve an inflammatory process followed by fibrosis.

Natural history

The review suggests symptoms peak and improve over a 1-3y period. It usually goes through 4 distinct phases though may develop more rapidly.

1 (Painful)0-3 mGradual onset of pain – present at night – no restriction of movement. Diagnosis tricky at this stage. Consider also referred pain from neck/lung.
2 (Freezing)3-9 mPain remains but now gradual loss of movement in all directions – starts to impact on daily activities.
3 (Frozen)9-15 mPain reduces – now significant restriction of active and passive movement in all directions.
4 (Thawing)12 m+Pain ceases and movement gradually restored over a 6m period.


  • Diagnosis is clinical-based on sequence of pain and stiffness described above. There are no specific diagnostic tests.
  • Plain X-ray may be helpful in ruling out severe osteoarthritis and other conditions such as bony metastases and apical lung tumours.


Unfortunately, there is no perfect treatment and crucial to management is patient education and managing expectations about the likely natural history. After this, treatments can be focused on improving pain, increasing movement and reducing disability with various options available.

Treatment optionEvidence
Watchful waiting with patient educationBaseline intervention.
Oral analgesiaNo evidence of superiority – NICE CKS recommend paracetamol +/- codeine or NSAIDs.
PhysiotherapySmall RCT showed hospital group-based exercise classes were more effective than individual therapy which was superior to home exercises.
Electrotherapy (light/sound/thermal)Short-term benefits in pain and function compared to placebo and home exercises.
Intra-articular corticosteroidShort-term benefits in pain and function and disability compared to placebo in first 6m.
Sodium hyaluronate injectionsNot recommended: no difference in pain and function or disability compared with a single corticosteroid injection (more expensive and 3 injections needed).
Hydrodistension (injecting ≥20ml saline into glenohumeral joint until capsule ruptures)Short-term benefits in pain and function and range of movement compared with placebo – but no additional benefits compared with home exercises or steroid injection.
Suprascapular nerve blockNo good quality RCTs.
Manipulation/arthroscopic release under GA.No good quality RCTs (UK-based RCT currently recruiting).


Consider referral in these situations:

  • No response to non-operative treatments after 6m.
  • Considering opiate analgesia.
  • Diagnostic uncertainty.
  • Earlier referral where the condition is particularly disabling, e.g. athletes, manual labourers.

Patient resources

A useful information leaflet can be found here.

No Comments Yet.

Leave your comments