This is a double-blinded study of subacromial corticosteroid injection (steroid injection to the shoulder) to treat shoulder pain in the paralyzed (hemiplegic) shoulder of chronic stroke survivors. This study is designed to evaluate pain relief of a standard steroid injection treatment, compared to a high dose treatment and a low dose treatment, for shoulder pain in stroke survivors. A total of 105 chronic stroke survivors with moderate to severe shoulder pain will be enrolled. All eligible participants will undergo an initial test injection to localize pain to the subacromial space. If this turns out to be positive, the subjects will be randomly assigned to one of three groups:

1. low dose group which receives 20mg of steroid (triamcinolone) injection to the subacromial space of the affected shoulder;
2. standard dose group which receives a standard 40mg of steroid (triamcinolone) injection to the subacromial space of the affected shoulder; or
3. high dose group which receives 60mg of steroid (triamcinolone) injection to the subacromial space of the affected shoulder.

Study participants will all rate their pain in interviews (Baseline, weeks 1, 2, 3, 4, 8, 12 (7 times) and in laboratory-based measures that will be administered at baseline, weeks 4, 8, 12 (4 times). Subjects will be followed for a total of 13 weeks.

The study will thus characterize the dose response of triamcinolone for the treatment of hemiplegic shoulder pain.

• A total of 105 chronic stroke survivors with moderate to severe shoulder pain will be enrolled. Subjects will be enrolled in the study for 13 weeks. Subjects who complete the protocol will visit MetroHealth five times.
• Visit 1: Baseline information about demographics, past medical history, and inclusion/exclusion issues will be collected for study participants. Pertinent lab work will be performed to determine initial eligibility.
• Visit 2: All initially eligible participants also will undergo a test injection of lidocaine to localize the pain to the subacromial bursa (Neer's Test). A positive Neer's test is required to finalize eligibility for further participation. Participants who satisfy inclusion/exclusion criteria (including a positive Neer's test) will be randomized to high dose steroid, standard dose steroid, or low dose steroid via a computer generated random number table. The study participants and the observer will be blinded as to these groupings.
• After the initial Neer's test and randomization, participants will receive their assigned injection the same day. Participants then will be followed for an additional 12 weeks, including three follow-up visits (Visits 3-5). The total participation time in this study will be 13 weeks.
• The primary outcome measure will be the BPI 12. The BPI is a pain questionnaire, which assesses the "worst pain" in the previous 7 days. Secondary outcome measures also will be assessed together with BPI 12. There will be 3 additional secondary outcome measures, Fugl-Meyer Motor Assessment (a measure of poststroke motor impairment), pain free external rotation range of motion (ROM) and pain free abduction ROM.
• A blinded therapist will administer all outcome measures. The primary outcome will be assessed on a weekly basis via telephone (or in person during weeks of MetroHealth visits) starting on the day of Visit 1 and continuing to 12-weeks after steroid injection (i.e., for the 13 weeks of the subject's participation). The remaining secondary outcomes will be assessed in clinic visits at least every 4 weeks starting with Visit 2 (Visits 2-5).

Inclusion Criteria:

• Age greater than 18
• upper extremity hemiplegia due to hemorrhagic or nonhemorrhagic stroke
• ≤ 4/5 on manual muscle testing for the deltoid on the affected side if isolated movement is present
• post-stroke duration ≥ 1-mo, but < 24-mo
• shoulder pain sustained for ≥ 1-mo
• BPI 12 ≥ 4 (pain scale)
• willing and able to report pain and other conditions throughout the 4-mo study period
• positive Neer's test

Exclusion Criteria:

• evidence of joint or overlying skin infection
• > 2 opioid and/or nonopioid analgesic for shoulder pain (i.e., > 2 regardless of class)
• regular intake of pain medications for any other chronic pain
• steroid injections to the shoulder in the last 6-wks
• history of pre-stroke shoulder pain
• bleeding disorder
• for those on Coumadin, INR > 3.0
• history of allergies to lidocaine
• renal insufficiency (Creat > 2.0)
• both history of liver disease & abnormal liver enzyme lab results
• poorly controlled diabetes (HbA1c > 7.0)
• medical instability
• cognitive deficits; In order to pass the cognitive exam, the participant must exhibit 3/3 registration and recall the same three objects in 30 minutes. The participant must also be able to correctly rank the following three levels of pain from highest to lowest: 1) falling from a 2-story building and breaking both ankles, 2) stubbing one's toe and 3) getting bitten by a mosquito.
• immunocompromised

The number of participants was determined on a power analysis. The goal was 31 subjects per group.

The study was powered to detect differnces between high-dose and placebo, and standard-dose and placebo. The trial was altered due to ethical concerns for placebo injection due to apparent effectiveness in the literature. Low dose was substituted.