CLINICAL QUESTION

Do intra-articular steroid injections improve pain and function in patients with adhesive capsulitis, also known as frozen shoulder?

EVIDENCE-BASED ANSWER

Intra-articular steroid injections inconsistently improve pain and function when used to treat adhesive capsulitis in the short term (less than 12 weeks) and medium term (less than 6 months). (Strength of Recommendation [SOR]: A, multiple meta-analyses, systematic review.) Ultrasound guidance improves accuracy of injection into the glenohumeral joint, but there is no difference in pain at 3, 6, and 12 weeks compared with landmark-guided injection. (SOR: B, single randomized controlled trial [RCT].)

EVIDENCE SUMMARY

A 2020 systematic review and meta-analysis of 65 randomized trials compared the relative effectiveness of nonsurgical treatments for adhesive capsulitis.¹ The systematic review included studies that were randomized designs of any type comparing treatment modalities with placebo injection, no treatment, or other treatment modalities. Primary outcomes were pain and function, and the secondary outcome was shoulder external rotation range of motion (ROM). Pain was evaluated using a 0- to 10-point visual analog scale (VAS), with lower numbers indicating less pain. Because the studies used different functional scores, standardized mean differences were used in the overall analysis. Shoulder external rotation ROM was measured in degrees. Minimal clinically important differences were defined as 1 point for VAS pain, an effect size of 0.45 for functional score, and 10 degrees for ROM.

Of the 65 eligible studies, 34 studies (n = 2,402) were included in a pairwise meta-analysis that compared the effectiveness of each intervention with other interventions or placebo injection/no treatment in the early short term (2 to 6 weeks), late short term (8 to 12 weeks), and medium term (4 to 6 months). Data beyond these periods were inadequate for analysis. Intra-articular steroid injections did not improve pain after 4 to 6 months. Significant improvement in function was present at 6 weeks but not at 6 months. The statistically significant benefits did not consistently reach the threshold of minimal clinical importance.

A limitation of the systematic review included grouping patients with various chronicities of adhesive capsulitis together, making it difficult to draw conclusions about treatment at specific presentation time frames. Also, the placebo-injection and no-treatment groups were combined, minimizing the significance of injection placebo effect.

A 2019 meta-analysis compared intra-articular and subacromial steroid injections for adhesive capsulitis.² Seven RCTs (N = 421) were included in the study. Inclusion criteria were studies of patients older than 18 years with adhesive capsulitis and intervention comparators of intra-articular and subacromial injection. Outcomes were pain measured via a 0- to 10-point VAS, ROM in degrees, and Constant-Murley score. This score is calculated using a 100-point composite scale divided into four categories: pain (15 points), activities of daily living (20 points), strength (25 points), and ROM (i.e., forward flexion, external rotation, internal rotation, and abduction; 40 points). A higher score reflects a greater level of function. Pain outcomes were reported at 1, 2, and 3 months after the interventions. ROM and Constant-Murley score data were recorded at 1 and 2 months.

Intra-articular steroid injection significantly improved pain compared with subacromial injection at 1 month (−0.87 VAS; 95% CI, −1.69 to −0.05), 2 months (−0.84 VAS; 95% CI, −1.58 to −0.09), and 3 months (−0.58 VAS; 95% CI, −1.07 to −0.09). ROM was not significantly different at 1 month (2.26; 95% CI, −2.28 to 6.80) or 2 months (−0.68; 95% CI, −4.98 to 3.62). Intra-articular and subacromial injections showed no difference in Constant-Murley scores at 1 month (3.95; 95% CI, −1.07 to 8.97) or 2 months (10.16; 95% CI, −3.72 to 24.04), but there was an overall improvement in the score using combined time-point data (7.13; 95% CI, 1.76 to 12.49).

Limitations of this study included small sample size and only short- to medium-term follow-up. Although there was a statistically significant difference in VAS pain scores, this less than 10% difference likely was not a minimal clinically important difference.

A 2021 prospective, double-blind, randomized trial investigated ultrasound-guided vs. blind or anatomic landmark– guided injection for adhesive capsulitis.³ Ninety patients with primary frozen shoulder were randomly assigned to ultrasound-guided (n = 45) or landmark-guided injection (n = 45) by an orthopedic shoulder specialist. Accuracy of the injections was determined using fluoroscopic imaging. Additional outcomes assessed were VAS for pain, American Shoulder and Elbow Surgeons (ASES) score, subjective shoulder value, and ROM at the time of presentation and at 3, 6, and 12 weeks postinjection. The ASES score assesses the condition of the shoulder using a survey with sections on pain and activities of daily living. The score ranges from 0 to 100 points, with a higher score indicating a shoulder in better condition. The subjective shoulder value is defined as the patient's subjective shoulder assessment expressed as a percentage of an “entirely normal” shoulder, with 100% representing a normal shoulder.

Both the ultrasound- and landmark-guided injection groups had significant improvement in VAS pain scores, ASES scores, subjective shoulder value, forward flexion, abduction, and external and internal rotation through 12-week follow-up compared with baseline. The accuracy of injection in the ultrasound-guided injection group was 100% compared with 71.1% in the landmark-guided injection group (P < .001). Despite the greater accuracy of ultrasound-guided injections, there were no differences in any pain or functional outcome measures compared with landmark-guided injections. No adverse events were noted in either group. The small sample size was a study limitation. The injections were performed by shoulder specialists, which may limit the generalizability to all clinicians.

RECOMMENDATIONS FROM OTHERS

A 2013 adhesive capsulitis clinical practice guideline from the American Physical Therapy Association states that intraarticular steroid injection combined with physiotherapy (i.e., shoulder mobility and stretching exercises) is more effective in providing short-term pain relief and improves function compared with physiotherapy alone.⁴