This is a study on the treatment of frozen shoulder with injection of cortisone into the shoulder joint. The purpose of the study is to evaluate whether injection with stretching of the shoulder capsule or plain injection is more effective than no specific treatment.

Shoulder capsulitis or frozen shoulder is a painful condition causing limitations of shoulder movements. Its incidence is about 2%, comprising mostly women. There is a need for more controlled trials on frozen shoulder preferably in general practice, without using technical guiding instruments for injections as MRI or computer tomography, because the condition is routinely treated in general practice. It was therefore important to conduct this study as similar as possible to current practice. The procedure should be simple, safe and less time consuming.

Patients will be allocated to the study groups by computerized block-randomization with three permutations per block. Group 1 consists of patients receiving Triamcinolone 20 mg intraarticular injection with Lidocaine 10mg/ml 3 ml and a total of 4 ml solution. Group 2 patients will receive intraarticular Triamcinolone 20 mg, 3 ml Lidocaine and the rest with physiological natrium chloride 9 mg/ml, comprising a total volume from 8 ml and upwards up to 20 ml. Group 3 will serve as a control group and patients in this group could receive any other treatment other than corticosteroid injections or per oral corticosteroid medication. The control group will remain without treatment with corticosteroids, in injection or tablet form till 61 days, which is also the last day of the outcome measurements. The time interval between the 1st and 2nd treatment will be 7 days, between the 2nd and 3rd treatment 10 days, and between the 3rd and 4th treatment 14 days. Patients will also receive a follow-up by mail 6 and 12 months after the inclusion, where pain and function (SPADI) is investigated.

The patients will be explained and instructed in filling out schema for SPADI, NPRS, pain figure, for sleep Bergen Insomnia Scale (BIS), EPQ-N to measure neuroticism, and Subjective Health Complaints (SHC).

Inclusion Criteria:

• Patients who on clinical examination show reduced shoulder movements with pain where commonly lateral rotation is most restricted, medial rotation least restricted and abduction is somewhat restricted, somewhere in between the two other movements, is called a capsular pattern of movement confirming shoulder capsulitis
• Had had the condition for at least 2 months
• Reduction of movement is at least 30% of normal in two of the three passive movements
• Are ready to undergo treatment as set up before hand (agree to participate in one of the three groups)
• Agree not to undergo other treatment simultaneously directed at the stiff shoulder other than in the control group
• Patient has voluntarily given his/her written consent for participation
• Has not already received intraarticular corticosteroid injection prior to entry in the study less than 4 weeks earlier

Exclusion Criteria:

• Diabetes, as in some diabetic patients, steroids can disturb the regulation of their blood sugar levels.
• One of the two shoulder movements are normal, either passive abduction or passive lateral rotation
• Have some systemic disease where use of corticosteroids is either contraindicated or affect the coexisting disorder
• Patients with asthma using steroid inhalations or steroids orally
• There are other coexisting disorders in the arm or patient has a painful neck that can disturb pain and functional assessment of the shoulder
• Pregnant and breast feeding mothers
• Patients under 18 years of age
• Patients who do not understand Norwegian
• Patients who have reduced intellectual capacity for consent-giving