Effectiveness of Ultrasound-Guided Versus Anatomic Landmark–Guided Corticosteroid Injection on Pain, Physical Function, and Safety in Patients With Subacromial Impingement Syndrome: A Systematic Review and Meta-analysis : American Journal of Physical Medicine & Rehabilitation

Secondary Logo

Journal Logo

Original Research Articles

Effectiveness of Ultrasound-Guided Versus Anatomic Landmark–Guided Corticosteroid Injection on Pain, Physical Function, and Safety in Patients With Subacromial Impingement Syndrome

A Systematic Review and Meta-analysis

Deng, Xiaoyan MD; Zhu, Siyi MD; Li, Daishun MD; Luo, Yi MD; Zhang, Xin MD; Tan, Yanling MD; Li, Juan MD; He, Xia MD

Author Information
American Journal of Physical Medicine & Rehabilitation: December 2022 - Volume 101 - Issue 12 - p 1087-1098
doi: 10.1097/PHM.0000000000001940

Abstract

What Is Known

  • Previous systematic reviews yielded contradictory results regarding the efficacy and safety of ultrasound-guided (USG) versus anatomic landmark–guided (ALG) corticosteroid injections for the treatment of subacromial impingement syndrome (SIS).

What Is New

  • This meta-analysis included the largest sample size of participants to date and compared the most comprehensive set of outcome measures stratified by age, follow-up duration, and type of corticosteroid. We found that the USG injection offers significantly better clinical improvement than ALG corticosteroid injection for the treatment of SIS, with acceptable safety.

BACKGROUND

Subacromial impingement syndrome (SIS) is a of common cause of shoulder pain1,2 and accounts for 44%–65% of all shoulder complaints3 in people older than 40 yrs.4 The pathophysiological mechanism of SIS is the mechanical impingement of the rotator cuff tendons as they pass through the subacromial space, which leads to a condition of inflammation and irritation.5,6 Shoulder pain occurs when patients elevate their arms at or above the shoulder level,7 is usually confined to the anterolateral shoulder, and radiates to the medial and lateral sides of the humerus.8 Physical function and quality of life can be severely affected by shoulder pain.9

The primary aims of SIS treatment are to resolve the mechanical dysfunction and relieve pain.10 Corticosteroid injection into the subacromial space has been recommended as a standard treatment approach in the management of SIS.11 Traditionally, corticosteroid injections are administered under the guidance of anatomical landmarks.12 Henkus et al.13 reported that 30%–80% of shoulder girdle injections reaching the subacromial capsule were regarded as blind injections. Injections performed by image guidance (fluoroscopy or ultrasonography) have been shown to improve the accuracy of shoulder girdle injections.14 With recent advances in ultrasound imaging, ultrasound-guided (USG) injections have gained popularity and have been demonstrated to have improved efficacy in the treatment of SIS.14

However, despite improvements in the accuracy of injection placement by USG, the current evidence on the effect of USG versus ALG corticosteroid injections for the management of SIS is conflicting. Several studies15,16 indicated that USG injections for SIS resulted in significant improvements in pain relief and physical function compared with ALG injections, whereas others17,18 reported no positive results. The results from previous systematic reviews were contradictory and previous studies were limited by small sample sizes, ambiguous definitions of SIS, and increased costs were raised to justify the advantages of using USG injections for SIS.19–23 Therefore, to address these concerns, a systematic review and meta-analysis were performed to summarize and update the current evidence on the efficacy and safety of applying USG versus ALG corticosteroid injection in the management of SIS.

METHODS

This systematic review and meta-analysis were conducted in accordance with the recommendations of the Cochrane Collaboration24 and reported based on the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines (see Supplementary Material Appendix A, Supplemental Digital Content 1, https://links.lww.com/PHM/B452).25 The protocol of this study is available in PROSPERO (CRD42020162682).

Search Strategy

The following electronic databases were searched from inception to August 15, 2021; using search terms shoulder impingement syndrome/impingement and corticosteroid injection: PubMed, Embase, Web of Science, the Cochrane Central Register of Controlled Trials, Scopus, ClinicalTrials.gov, CBM, CNKI, and Wanfang databases. Detailed electronic search strategies are provided in the Supplementary Material Appendix B (Supplemental Digital Content 2, https://links.lww.com/PHM/B453). Open Grey (http://www.opengrey.eu/) was searched for the gray literature research. The reference lists of the included reviews and trials were also screened to identify potentially related studies.

Inclusion and Exclusion Criteria

Types of Studies

All randomized controlled trials comparing the efficacy of USG versus ALG corticosteroid injection in treating SIS were eligible for inclusion. Studies without full text or case reports were excluded. The language of the included randomized controlled trials was limited to English and Chinese.

Types of Participants

Adults participants (older than 18 yrs) with (1) a history of shoulder pain that worsened because of overhead activities and lying on the affected shoulder; (2) painful restriction of active flexion and/or abduction of the shoulder, or limitations in internal and external rotation; and (3) a positive Neer test, Hawkins test, Jobes test, painful arch, or external rotation resistance test1 were included. Causes of shoulder pain other than SIS were excluded.

Types of Interventions

The intervention was a USG corticosteroid injection, and the comparator was an ALG corticosteroid injection. The doses and types of corticosteroids used were not limited. There was no restriction on the injection approach: anterior, lateral, and posterior approaches.

Types of Outcome Measures

The primary outcome was pain and secondary outcomes were function/disability, range-of-motion (ROM) assessment, and adverse events. The outcomes were prioritized as suggested by Steuri et al.26 and pain or functional outcomes were assessed by using different scales (Supplementary Material Appendix C, Supplemental Digital Content 3, https://links.lww.com/PHM/B454).

Study Selection

A three-stage screening methodology was used to select relevant randomized controlled trials for this review. First, all titles were screened by one reviewer (XYD) for eligibility, and irrelevant articles were excluded accordingly. Second, two reviewers (SYZ and DSL) independently reviewed the titles and abstracts of each study. Third, two independent reviewers (DSL and XYD) accessed the full text to assess against the eligibility criteria for each potentially eligible study. Disagreements were resolved through discussion with a third party (SYZ).

Data Collection

Two reviewers (XYD and SYZ) independently extracted the data from studies with a standardized spreadsheet, including data on lead author, year of publication, country, sample size, participants (sex and age), interventions, approach, follow-up period, and outcomes. Two reviewers cross-checked the extracted data. For missing data, we emailed the corresponding author or estimated the mean, SD, confidence interval (CI), or P values.24 All discrepancies were arbitrated by another author (YL).

Quality Assessment

We used the Cochrane Collaboration’s 2.0 tool (RoB 2)27 to appraise the risk of bias for eligible studies in the following domains: randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, selection of the reported results, and overall bias. The risk of bias in each domain was judged as “low,” “some concerns,” or “high.” The Grades of Recommendation Assessment, Development, and Evaluation (GRADE) tool28 was used to describe the overall quality of the body of evidence. It contains the following domains: risk of bias, inconsistency, indirectness, imprecision, and publication bias. The certainty of evidence was classified into four categories: high, moderate, low, or very low for each outcome. Two reviewers independently assessed the risk of bias for each study (XZ and DSL), and a third author (YLT) reviewed the final results and resolved any disagreements.

Statistical Analysis

All statistical analyses were performed using Revman (version 5.4) and STATA 15.0 software. Pooled effect assessments were analyzed by comparing changes between baseline and posttreatment outcomes: outcomeposttreatment and outcomebaseline. The changes in pain scores, function scores, and ROM were expressed as mean differences (MDs) with 95% CIs. If the evaluation scales differed for the same outcome, the standard MD (SMD) was calculated. To summarize the safety outcome (number of adverse events), the risk ratios with 95% CIs were used. We used the I2 value to detect the heterogeneity of the pooled results, as follows: I2 > 50% was defined as significant heterogeneity; I2 < 50% was defined as no significant heterogeneity. A fixed-effects model was used to combine studies with I2 < 50%. Otherwise, random-effects models were used. Subgroup analysis was conducted to detect the effectiveness relative to different age (≥50 and <50 yrs old), follow-up durations (1–2, 4, 6–8, and 12 wks), and different types of corticosteroids (betamethasone, dexamethasone, methylprednisolone). To evaluate the quality and consistency of pooled results, a sensitivity analysis was performed by excluding studies one by one to determine whether the changes had a significant impact on the treatment effect. Publication bias was estimated using a funnel plot if the comparisons included at least 10 trials.29 Begg test,30 and Egger test,31 and trim and fill analyses were used for further quantitative analysis of publication bias. All tests were two-tailed, and statistical significance was set at P < 0.05.

RESULTS

Study Characteristics

A total of 298 studies were obtained. Of these, 296 studies were identified using the original databases, and two additional studies were identified from published systematic reviews.19,20 After deleting the duplicate articles and reviewing the titles and abstracts, 47 studies were selected. Among these, 12 studies that met the inclusion criteria for full-text screening were obtained for the data extraction and analysis (Fig. 1). A total of 891 participants (USG group, n = 454; ALG group, n = 437) were enrolled. The sample size of individual studies ranged from 14 to 128 participants per group with a mean age between 28.13 and 57.7 yrs. The average disease duration ranged from 1 to 26 mos. The follow-up duration of the included trials ranged from 1 to 12 wks. The injection approach varied across the studies. The characteristics of the included studies are listed in Table 1.

F1
FIGURE 1:
Study flow diagram.
TABLE 1 - Characteristics of included studies
Author (Year), Country No. Patients Women,
n/%
Mean Age (SD)/Media (Min–Max) Type of Intervention and Dose Approach Duration, mo/wk Follow-up Outcomes Measure
Chen 32 (2006), Taiwan 20 26/33 53 (30–66) USG: 1 ml of betamethasone and 1 ml of 1% lidocaine Lateral >1 mo 1 wk Abduction ROM (not specified active or passive)
20 ALG: 1 ml of betamethasone and 1 ml of 1% lidocaine Lateral
Panditaratne 33 (2010), United Kingdom 41 36/62 54 (25–80) USG: 80 mg of Depo-Medrol and 3–10 ml of bupivacaine or lidocaine NR NR 6 wks Pain (not specified active or rest)
The Oxford shoulder score
17 ALG: 80 mg of Depo-Medrol and 3–10 ml of bupivacaine or lidocaine Lateral
Dogu 34 (2012), Turkey 23 15/65.2 55.17 (9.24) USG: 1 ml of 5 mg/ml of betamethasone dipropionate and 9 ml of 10 mg/ml prilocaine hydrochloride Lateral 7.43 (5.37) mos 6 wks VAS (rest, activity and sleep)
the constant ROM score
Constant general pain
Constant activities of daily living score
SDQ
23 16/69.6 56.74 (8.02) ALG: 1 ml of 5 mg/ml betamethasone dipropionate and 9 ml of 10 mg/ml prilocaine hydrochloride Posterior 9.74 (7.67) mos
Hsieh 16 (2013), Taiwan 48 27/56 57.59 (10.30) USG: 0.5 ml (5 mg·ml−1) of dexamethasone and 3 ml (10 mg·ml−1) of lidocaine hydrochloride Lateral 6.28 (3.59) mos Immediately
1 wk
4 wks
VAS (not specified active or rest)
SPADI
SDQ
SF-36
ROM of flexion, abduction,  external rotation, internal rotation (active and passive)
48 29/60 55.87 (11.42) ALG: 0.5 ml (5 mg·ml−1) of dexamethasone and 3 ml (10 mg·ml−1) of lidocaine hydrochloride Lateral 7.14 (4.72) mos
Saeed 35 (2014), Ireland 59 79/65 57.7 USG: 40 mg of methylprednisolone acetate with 4 ml of lidocaine hydrochloride NR 19.64 (1.84) wks 6 wks
12 wks
VAS (not specified active or rest)
SFTs
PGA
66 ALG: 40 mg of methylprednisolone acetate with 4 ml of lidocaine hydrochloride Lateral 20.02 (1.52) wks
Haghighat 15 (2015), Iran 20 12/60 50.45 (6.78) USG: 40 mg of methylprednisolone with 1 ml of lidocaine 2% Lateral 1.8 (0.54) mos 6 wks VAS (not specified active or rest)
SPADI
ROM of flexion, abduction, external  rotation, internal rotation (not specified if active or passive)
20 13/65 52.3 (7.48) ALG: 40 mg of methylprednisolone with 1 ml of lidocaine 2% Posterior 1.87 (0.48) mos
Cole 36 (2016), Australia 28 14/ (50) 46 (19–68) USG: 1 ml of 40 mg/ml methylprednisolone acetate and 5 ml of 1% lidocaine hydrochloride Lateral 26 (1–108) mos 6 wks VAS (active)
ASES
28 18/ (64) 42 (23–62) ALG: 1 ml of 40 mg/ml methylprednisolone acetate and 5 ml of 1% lidocaine hydrochloride Posterior 16 (2–108) mos
Bhayana 18 (2018), India 30 17/56.6 44.53 (9.2) USG: 2 ml of 40 mg/ml methylprednisolone acetate suspension mixed and 2 ml of 1% lignocaine Anterior 2 mos Day 5
3 wks
6 wks
12 wks
VAS (not specified active or rest)
Constant score
30 10/33.3 42.03 (9.9) ALG: 2 ml of 40 mg/ml methylprednisolone acetate suspension mixed and 2 ml of 1% lignocaine Lateral
Cao 37 (2018), China 28 7/25 50.2 (3.6) USG: 1 ml of betamethasone and 1 ml of 2% lidocaine Lateral 18 (4) mos 4 wks CMS
ADL
ROM
28 9/25 47.6 (2.6) ALG: 1 ml of betamethasone and 1 ml of 2% lidocaine Lateral 16 (5) mos
Akbari 17 (2020), Turkey 14 8/57.1 39.5 (20–64) USG: methylprednisolone acetate 40 mg in 1 ml and procaine 2% 4 ml Anterior >3 mos 4 wks VAS (not specified active or rest)
Flexion, abduction ROM (active)
DASH
CMS
14 9/64.3 42.5 (20–64) ALG: methylprednisolone acetate 40 mg in 1 ml and procaine 2% 4 ml Posterior
Roddy 38 (2021), United Kingdom 128 63/51 53.8 (10.2) USG: methylprednisolone 40 mg and 1 ml of 1% lidocaine Anterior Time varied 6 wks
6 mos
12 mos
NRS
SPADI
Patient’s Self-reported Global  Impression of Change
Short-Form 12
128 67/128 53.8 (10.2) ALG: methylprednisolone 40 mg and 1 ml of 1% lidocaine Lateral
Azadvari 39 (2021), Iran 15 10/66.7 33.80 (9.34) USG: 1 ml of Depo-Medrol along
with 2 ml of lidocaine
Lateral 23.46 (14.78) mos 2 wks
2 mos
CONSTANT score
VAS
BREF-QOL questionnaire
15 12/80 28.13 (11.02) ALG: 1 ml of Depo-Medrol along
with 2 ml of lidocaine
Lateral 22.53 (14.87) mos
ADL, activities of daily living; ASES, American Shoulder and Elbow Surgeons score; BREQ, Brief quality of life; CMS, modified Constant-Murley Score; DASH, Disabilities of the Arm, Shoulder, and Hand questionnaire; NR, not reported; NRS, Numerical Rating Scale; PGA, Physician Global Assessment; QOL, quality of life; SDQ, Shoulder Disability Questionnaire; SF-36, 36-item Short-Form Health Survey; SFTs, shoulder function tests; SPADI, Shoulder Pain and Disability Index.

Pain data were obtained from 10 studies15–18,33–36,38,39 with the use of Visual Analog Scale (VAS) and Numerical Rating Scale, and only one study34 reported the type of pain (active and passive). Functional and disability outcomes were assessed in 11 trials15–18,33–39 using different tools (shoulder pain and disability index, modified Constant-Murley score, Oxford shoulder score, shoulder disability questionnaire, physician global assessment, American shoulder, and elbow surgeons score). Four studies15–17,32 reported ROM assessment. Adverse events were reported in six studies.17,18,34,35,37,38

Quality Assessment

Of the 12 included studies, one was rated as having a “low risk of bias” in all domains, and the other studies were classified as having an “unclear risk of bias” for at least one aspect or a “high risk of bias” for at least two aspects. The results of the risk of bias are presented in the Supplementary Material Appendix D (Supplemental Digital Content 4, https://links.lww.com/PHM/B455). According to the GRADE system, the quality of the evidence of the included studies was very low, low, or moderate. For pain, the certainty of the evidence was moderate. The evidence was downgraded because of the risk of bias. For function, the certainty of the evidence was very low. The evidence was downgraded because of the risk of bias, inconsistency, and imprecision. For shoulder ROM, the certainty of the evidence was low or very low. The evidence was downgraded because of the risk of bias, inconsistency, and imprecision. The GRADE results are presented in Table 2.

TABLE 2 - Results of the GRADE
Certainty Assessment No. Patients Effect Certainty
Outcomes No. Studies Study Design Risk of Bias Inconsistency Indirectness Imprecision Publication Bias USG Injection ALG Injection Absolute (95% CI)
Pain 10 Randomized trials Serious a Not serious Not serious Not serious None 406 389 MD = 0.58 lower
(1.05 lower to 0.1 lower)
⨁⨁⨁◯
Moderate
Function 11 Randomized trials Serious a Serious b Not serious Serious c Publication bias strongly suspected d 434 417 SMD = 0.82 SD lower
(1.38 lower to 0.26 lower)
⨁◯◯◯
Very low
Abduction of ROM 4 Randomized trials Serious a Serious b Not serious Not serious None 102 102 MD = 9.28 higher
(12.13 lower to 30.7 higher)
⨁⨁◯◯
Low
Flexion of ROM 3 Randomized trials Serious a Serious b Not serious Serious e None 82 82 MD = 1.53 lower
(15.54 lower to 12.47 higher)
⨁◯◯◯
Very low
External rotation of ROM 2 Randomized trials Serious f Not serious Not serious Serious g None 68 68 MD = 0.28 lower
(1.38 lower to 0.38 higher)
⨁⨁◯◯
Low
Internal rotation of ROM 2 Randomized trials Serious f Not serious Not serious Serious g None 68 68 0.77 lower
(1.72 lower to 0.18 higher)
⨁⨁◯◯
Low
a Two trials did not describe the randomization process; in one trial, patients and the physician administering the injections were not blinded.
b I2 value of the combined results was high heterogeneity.
c Potential imprecision due to different scales was combined for analysis for functional performance.
d There was a suspicion of publishing bias.
e Potential imprecision due to availability of three trials of 164 participants.
f Patients and the physician administering the injections were not blinded.
g Potential imprecision due to availability of two trials of 136 participants.

Effect of Intervention: Pain Relief

Ten trials15–18,33–36,38,39 (795 participants) were included in the analysis. Moderate certainty evidence showed that the USG injection had a small but significant effect on the reduction of pain (10 trials; MD = −0.58; 95% CI = −1.05 to −0.10; P = 0.017) with statistically significant heterogeneity (I2 = 76.7%, P = 0.000; Fig. 2). In the subgroup analysis, follow-up duration of 6–8 wks (8 trials; MD = −0.72; 95% CI = −1.19 to −0.25; P = 0.002) and use of methylprednisolone (8 trials; MD = −0.79; 95% CI = −1.28 to −0.29; P = 0.002) and betamethasone (1 trial; MD = −0.94; 95% CI = −1.58 to −0.30; P = 0.004) showed evidence of a significant difference in pain relief (Table 3). Age and other subgroup factors had no effect on pain relief. No significant changes in heterogeneity or overall effects were observed in the sensitivity analysis (Supplementary Material Appendix E, Supplemental Digital Content 5, https://links.lww.com/PHM/B456).

F2
FIGURE 2:
Forest plot for effects of USG versus ALG injection on pain.
TABLE 3 - Subgroup analyses of USG versus ALG injection of corticosteroid for pain and function
Pain Function
Subgroup Articles Effect Size (95% CI) P Articles Effect Size (95% CI) P
Age, yr
 <50 4 −1.20 (−2.84 to 0.44) 0.15 4 −0.076 (−1.95 to 0.43) 0.21
 >50 6 −0.35 (−0.73 to 0.03) 0.07 6 −0.46 (−1.05 to 0.13) 0.13
Flow-ups
 1–2 wks 2 −1.51 (−5.22 to 2.21) 0.43 2 −1.35 (−4.72 to 2.01) 0.43
 4 wks 2 0.46 (−1.29 to 2.20) 0.61 3 −0.85 (−2.81 to 1.11) 0.39
 6–8 wks 8 −0.72 (−1.19 to −0.25) 0.002 8 −0.78 (−1.36 to −0.21) 0.008
 12 wks 2 −0.53 (−1.25 to 0.20) 0.15 2 −0.23 (−0.66 to 0.21) 0.31
Drugs
 Dexamethasone 1 0.86 (−0.03 to 1.75) 0.06 1 0.49 (0.08 to 0.90) 0.02
 Betamethasone 1 −0.94 (−1.58 to −0.30) 0.004 2 −1.61 (−4.07 to 0.84) 0.2
 Methylprednisolone 8 −0.79 (−1.28 to −0.29) 0.002 8 −0.77 (−1.36 to −0.18) 0.01

Effect of Intervention: Function Improvement

Eleven trials15–18,33–39 (851 participants) were included in the analysis. Very low certainty evidence showed that USG injection had a small but significant effect on the improvement of function (11 trials; SMD = −0.84; 95% CI = −1.41 to −0.27; P = 0.004) with statistically significant heterogeneity (I2 = 92.8%, P = 0.000; Fig. 3). In the subgroup analysis, follow-up duration of 6–8 wks (8 trials; SMD = −0.78; 95% CI = −1.36 to −0.21; P = 0.008) and use of methylprednisolone (8 trials; SMD = −0.77; 95% CI = −1.36 to −0.18; P = 0.01) showed evidence of a significant difference in functional improvement (Table 3). Age and other subgroup factors had no effect on functional improvement. The meta-analysis results for function were robust in sensitivity analysis (Supplementary Material Appendix E, Supplemental Digital Content 5, https://links.lww.com/PHM/B456).

F3
FIGURE 3:
Forest plot for effects of USG versus ALG injection on function.

Effect of Intervention: ROM

Four trials15–17,32 (204 participants) were included in the analysis. Very low to low certainty evidence showed that USG injection had no significant effect on the improvement of shoulder ROM (Fig. 4). In the subgroup analysis, follow-up duration of 6 wks (1 trial; MD = 6.50; 95% CI = 4.50 to 8.50; P < 0.00001) and use of betamethasone (1 trial; MD = 41.54; 95% CI = 30.95 to 52.13; P < 0.00001) showed evidence of a significant difference in shoulder abduction ROM. Age 50 yrs or greater (2 trials; MD = 4.51; 95% CI = 3.29 to 5.74; P < 0.0001) and follow-up duration of 6 wks (1 trial; MD = 4.50; 95% CI = 3.27 to 5.37; P < 0.0001) showed evidence of a significant difference in shoulder flexion ROM (Table 4). Other subgroup factors had no effect on shoulder ROM. The results showed no major change in the overall findings in the sensitivity analysis, which suggested the stability of the results (Supplementary Material Appendix E, Supplemental Digital Content 5, https://links.lww.com/PHM/B456).

F4
FIGURE 4:
Forest plot for effects of USG versus ALG injection on ROM. A, Shoulder abduction ROM. B, Shoulder flexion ROM. C, Shoulder external rotation ROM. D, Shoulder internal rotation ROM.
TABLE 4 - Subgroup analyses of USG versus ALG injection of corticosteroid for ROM
ROM of Abduction ROM of Flexion
Subgroup Studies Effect Size (95% CI) P Studies Effect Size (95% CI) P
Age, yr
 <50 1 −10.00 (−28.69 to 8.69) 0.29 1 −20.00 (−36.72 to −3.28) 0.02
 ≥50 3 15.03 (−11.46 to 41.52) 0.27 2 4.51 (3.29 to 5.74) <0.0001
Flow-ups
 1 wk 2 21.59 (−19.99 to 63.17) 0.31 1 10.23 (−5.78 to 26.23) 0.21
 4 wks 2 −8.40 (−22.88 to 6.08) 0.26 2 −6.21 (−33.08 to 20.66) 0.65
 6 wks 1 6.50 (4.50 to 8.50) <0.00001 1 4.50 (3.27 to 5.73) <0.0001
Type of corticosteroid
 Dexamethasone 1 −6.01 (−28.89 to 16.87) 0.61 1 6.96 (−9.48 to 23.40) 0.41
 Betamethasone 1 41.54 (30.95 to 52.13) <0.00001
 Methylprednisolone 2 0.97 (−14.29 to 16.24) 0.90 2 −6.27 (−30.11 to 17.56) 0.61
ROM of External Rotation ROM of Internal Rotation
Subgroup Studies Effect Size (95% CI) P Studies Effect Size (95% CI) P
Age, yr
 <50 0 0
 ≥50 2 −0.28 (−1.38 to 0.83) 0.62 2 −0.77 (−1.72 to 0.18) 0.11
Flow-ups
 1 wk 1 −2.80 (−12.01 to 6.41) 0.55 1 −8.24 (−22.98 to 6.51) 0.27
 4 wks 1 −3.52 (−15.77 to 8.73) 0.57 1 −6.43 (−22.02 to 9.16) 0.42
 6 wks 1 −0.25 (−1.36 to 0.86) 0.66 1 −0.75 (−1.70 to 0.20) 0.12
Type of corticosteroid
 Dexamethasone 1 −3.52 (−15.77 to 8.73) 0.56 1 −6.43 (−22.02 to 9.16) 0.81
 Betamethasone
 Methylprednisolone 1 −0.25 (−1.36 to 0.86) 0.44 1 −0.75 (−1.70 to 0.20) 0.12

Safety Outcome

Six studies reported adverse events. Two studies35,38 reported that shoulder pain occurred in a small number of cases. In one study,38 a serious adverse event was observed in a participant who was hospitalized for pyelonephritis. Four studies reported that no adverse events were observed.17,18,34,37

Publication Bias

For pain, the funnel plot was symmetry and Begg test (z = 1.07, P = 0.283) and Egger (t = −0.49, P = 0.637) tests did not detect publication bias. For function, the funnel plot showed a mild asymmetry and Begg (z = 2.02, P = 0.043), Egger (t = −2.98, P = 0.024) tests indicated publication bias. Trim and filled analyses did not change, which showed that the results were stable. For ROM, the Begg and Egger tests showed no publication bias. The results are presented in Supplementary Material Appendices F1–F4 (Supplemental Digital Content 6, https://links.lww.com/PHM/B457).

DISCUSSION

This systematic review and meta-analysis of 12 studies involving 891 participants (USG group: 454 participants; ALG group: 437 participants) suggested that USG corticosteroid injection for management of SIS resulted in more effective pain relief and functional improvement than ALG injections; however, there was no significant difference in shoulder ROM. In the subgroup analysis, follow-up of 6–8 wks and use of methylprednisolone showed evidence of a significant difference in pain relief and functional improvement.

Our findings are in consistent with those of several previously published systematic reviews.20,40,41 Several mechanisms have been suggested to explain the significantly greater improvement in pain reduction and functional gain in the USG group. The main mechanism is the greater accuracy of USG versus ALG for all shoulder girdle joints, which led to improved efficacy in outcome measures of pain and function.41 Another important factor resulting in the improvement to consider is the increased patient comfort and less needle manipulation with USG injections, although it was not assessed across all studies.20 Ultrasound-guided injections also provide real-time monitoring during needle placement without any risk of radiation exposure. Furthermore, a short-term retention of the therapeutic efficacy may exist as reported by the findings in our subgroup analysis, and another systematic review measured the changes between 6-wk follow-up and baseline VAS and shoulder function scores.40 However, several factors should be considered when interpreting these positive findings such as the varied quality of included studies, inclusion of specific populations with small sample sizes, substantial heterogeneity across studies, and small clinical effect. The reasons for these weaknesses could be attributed to different periods of follow-up, heterogeneous pathologies of painful shoulder, inadequate patient blinding potentially causing bias, and placebo effect.14 In our study, although there was an evidence of substantial heterogeneity among the included trials, our sensitivity analysis detected no changes in the levels of heterogeneity and effect, and the results of publication bias confirmed the robustness.

In contrast, a recent Cochrane review and another systematic review did not detect improved efficacy or advantages of USG corticosteroid injection for the management of SIS.19,23 Bloom et al.19 included patients with various pathologies for shoulder pain (one of which was SIS), whereas Ayekoloye et al.23 included patients with SIS under the injection of the subacromial bursa only. However, these previous systematic reviews,19,23 which obtained negative results, included only 4–5 studies involved 234–290 patients with different causes of shoulder pain, whereas our study included 12 trials involving 891 participants with SIS. In addition, participants included in previous studies were in the chronic stage of SIS, which is less responsive to corticosteroid injections reported,42 and these patients received additional treatments with the corticosteroid injection, which may have influenced the outcome measures.43 Other notable limitations of the abovementioned Cochrane review include the use of the final pain and functional outcome MD between groups instead of the mean change difference and the fact that negative conclusion in terms of pain, function, shoulder range of motion, or safety was made mainly based on a single study.44

Strengths and Limitations

This review has several strengths. To our knowledge, this study had the largest sample size of participants (891 participants in 12 trials) with SIS and evaluated the treatment efficacy of USG versus ALG injections. Our study used the most comprehensive set of outcome measures including pain, function, ROM, and safety, which were stratified by age, follow-up duration, and the type of corticosteroid. These results will guide clinicians in making a decision regarding the use of USG injections for the management of SIS or similar musculoskeletal disorders. Furthermore, to optimize data gathering and to ensure that no study was missed in compliance with the study protocol, we sought the help of a librarian familiar with the development of searches and the mechanism of living systematic reviews. Finally, this study strictly adhered to the Preferred Reporting Items for Systematic Reviews and Meta-analyses checklist and was carefully performed according to guidelines to ensure the robustness of the findings.

Nevertheless, this review also has several limitations. First, the level of evidence assessed by the GRADE criteria ranged from moderate to very low, which suggests that the quality of studies included needs to be further improved by including studies with a larger sample size and well-designed methodology to eliminate any potential risk of bias. Moreover, there were remarkable variations in the measures of pain and functional outcomes used between studies; therefore, we used SMD for all functional measures in the analysis. Pain measures across studies were converted to a 10-point scale, where some unspecified pain outcomes were added. Finally, SIS definitions in the included studies varied because of a lack of well-defined diagnostic criteria for SIS,45 and the systematic search was limited to the English and Chinese languages, which may have resulted in some relevant trials being missed.

Implications

Ultrasound-guided injections are associated with several advantages including the lack of radiation exposure; therefore, this technique has gained widespread use in recent years, replacing the conventional ALG injection, also known as “blind” approach.46 In our study, few cases of adverse events were reported in the safety analysis, although only six studies reported this outcome. Therefore, based on our findings, we recommend USG corticosteroid injection as an efficient and safe intervention for patients with SIS, particularly when a physician or therapist is not well trained and proficient in using ALG injection, both of which can achieve similar accuracy and efficacy when administered by an experienced provider.41 Although some researchers have expressed concerns regarding the lack of efficacy of USG injections to justify their higher cost, a cost-effectiveness analysis has not been performed in current evidence to address these concerns.41 Until this analysis is conducted, the decision regarding the use of USG versus ALG injections should be primarily informed by evidence from currently available efficacy analysis.

CONCLUSIONS

The meta-analysis in this study provides moderate to very low evidence that USG corticosteroid injection for the management of SIS results in significant pain relief and improvement in physical function of the shoulder as compared with ALG injection. Short-term retention between 6 and 8 wks after injection and the type of corticosteroid used potentially affect the treatment efficacy of USG as reported in this study. Adverse events were relatively low in both groups, which justifies the safety and efficacy of USG injection in treating SIS. Future research should optimize the study design (taking factors not limited to a clear definition of SIS), perform a cost-effective analysis, include longer follow-up periods and larger sample size, and perform an intention-to-treat analysis to generate high certainty evidence.

ACKNOWLEDGMENTS

The authors are grateful for the partnership and support from the library of Chengdu University of Traditional Chinese Medicine for the development of search strategy.

REFERENCES

1. Garving C, Jakob S, Bauer I, et al.: Impingement syndrome of the shoulder. Dtsch Arztebl Int 2017;114:765–76
2. Andersson HI: The course of non-malignant chronic pain: a 12-year follow-up of a cohort from the general population. Eur J Pain 2004;8:47–53
3. Michener LA, McClure PW, Karduna AR: Anatomical and biomechanical mechanisms of subacromial impingement syndrome. Clin Biomech (Bristol, Avon) 2003;18:369–79
4. Chang WK: Shoulder impingement syndrome. Phys Med Rehabil Clin N Am 2004;15:493–510
5. Bell AD, Conaway D: Corticosteroid injections for painful shoulders. Int J Clin Pract 2005;59:1178–86
6. Cools AM, Johansson FR, Borms D, et al.: Prevention of shoulder injuries in overhead athletes: a science-based approach. Braz J Phys Ther 2015;19:331–9
7. Creech JA, Silver S: Shoulder impingement syndrome, StatPearls. StatPearls Publishing, Copyright © 2021. Treasure Island, FL, StatPearls Publishing LLC., 2021
8. Karthikeyan S, Kwong HT, Upadhyay PK, et al.: A double-blind randomised controlled study comparing subacromial injection of tenoxicam or methylprednisolone in patients with subacromial impingement. J Bone Joint Surg Br 2010;92:77–82
9. MacDermid JC, Ramos J, Drosdowech D, et al.: The impact of rotator cuff pathology on isometric and isokinetic strength, function, and quality of life. J Shoulder Elbow Surg 2004;13:593–8
10. Dong W, Goost H, Lin XB, et al.: Treatments for shoulder impingement syndrome: a PRISMA systematic review and network meta-analysis. Medicine (Baltimore) 2015;94:e510
11. Gaujoux-Viala C, Dougados M, Gossec L: Efficacy and safety of steroid injections for shoulder and elbow tendonitis: a meta-analysis of randomised controlled trials. Ann Rheum Dis 2009;68:1843–9
12. Ombregt L: A System of Orthopaedic Medicine, 3rd ed. London, Churchill Livingstone, 2013
13. Henkus HE, Cobben LP, Coerkamp EG, et al.: The accuracy of subacromial injections: a prospective randomized magnetic resonance imaging study. Arthroscopy 2006;22:277–82
14. Hall MM: The accuracy and efficacy of palpation versus image-guided peripheral injections in sports medicine. Curr Sports Med Rep 2013;12:296–303
15. Haghighat S, Taheri P, Banimehdi M, et al.: Effectiveness of blind & ultrasound guided corticosteroid injection in impingement syndrome. Glob J Health Sci 2015;8:179–84
16. Hsieh LF, Hsu WC, Lin YJ, et al.: Is ultrasound-guided injection more effective in chronic subacromial bursitis?Med Sci Sports Exerc 2013;45:2205–13
17. Akbari N, Ozen S, Şenlikçi HB, et al.: Ultrasound-guided versus blind subacromial corticosteroid and local anesthetic injection in the treatment of subacromial impingement syndrome: a randomized study of efficacy. Jt Dis Relat Surg 2020;31:115–22
18. Bhayana H, Mishra P, Tandon A, et al.: Ultrasound guided versus landmark guided corticosteroid injection in patients with rotator cuff syndrome: randomised controlled trial. J Clin Orthop Trauma 2018;9:S80–5
19. Bloom JE, Rischin A, Johnston RV, et al.: Image-guided versus blind glucocorticoid injection for shoulder pain. Cochrane Database Syst Rev 2012:CD009147
20. Wu T, Song HX, Dong Y, et al.: Ultrasound-guided versus blind subacromial-subdeltoid bursa injection in adults with shoulder pain: a systematic review and meta-analysis. Semin Arthritis Rheum 2015;45:374–8
21. Sage W, Pickup L, Smith TO, et al.: The clinical and functional outcomes of ultrasound-guided vs landmark-guided injections for adults with shoulder pathology—a systematic review and meta-analysis. Rheumatology (Oxford) 2013;52:743–51
22. Soh E, Li W, Ong KO, et al.: Image-guided versus blind corticosteroid injections in adults with shoulder pain: a systematic review. BMC Musculoskelet Disord 2011;12:137
23. Ayekoloye CI, Nwangwu O: Ultrasound-guided versus anatomic landmark-guided steroid injection of the subacromial bursa in the management of subacromial impingement: a systematic review of randomised control studies. Indian J Orthop 2020;54(suppl 1):10–9
24. Cumpston M, Li T, Page MJ, et al.: Updated guidance for trusted systematic reviews: a new edition of the Cochrane Handbook for Systematic Reviews of Interventions. Cochrane Database Syst Rev 2019;10:Ed000142
25. Page MJ, McKenzie JE, Bossuyt PM, et al.: The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71
26. Steuri R, Sattelmayer M, Elsig S, et al.: Effectiveness of conservative interventions including exercise, manual therapy and medical management in adults with shoulder impingement: a systematic review and meta-analysis of RCTs. Br J Sports Med 2017;51:1340–7
27. Sterne JAC, Savović J, Page MJ, et al.: RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ 2019;366:l4898
28. Guyatt GH, Oxman AD, Vist GE, et al.: GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 2008;336:924–6
29. Higgins JP, Altman DG, Gøtzsche PC, et al.: The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ 2011;343:d5928
30. Begg CB, Mazumdar M: Operating characteristics of a rank correlation test for publication bias. Biometrics 1994;50:1088–101
31. Egger M, Davey Smith G, Schneider M, et al.: Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629–34
32. Chen MJ, Lew HL, Hsu TC, et al.: Ultrasound-guided shoulder injections in the treatment of subacromial bursitis. Am J Phys Med Rehabil 2006;85:31–5
33. Panditaratne N, Wilkinson C, Groves C, et al.: Subacromial impingement syndrome: a prospective comparison of ultrasound-guided versus unguided injection techniques. Ultrasound 2010;18:176–81
34. Dogu B, Yucel SD, Sag SY, et al.: Blind or ultrasound-guided corticosteroid injections and short-term response in subacromial impingement syndrome: a randomized, double-blind, prospective study. Am J Phys Med Rehabil 2012;91:658–65
35. Saeed A, Khan M, Morrissey S, et al.: Impact of outpatient clinic ultrasound imaging in the diagnosis and treatment for shoulder impingement: a randomized prospective study. Rheumatol Int 2014;34:503–9
36. Cole BF, Peters KS, Hackett L, et al.: Ultrasound-guided versus blind subacromial corticosteroid injections for subacromial impingement syndrome: a randomized, double-blind clinical trial. Am J Sports Med 2016;44:702–7
37. Cao W, Wang S, Yang X: Clinical study of ultrasound-guided acromial glide capsule injection for subacromial impingement syndrome. J Pract Pain 2018;4:272–6
38. Roddy E, Ogollah RO, Oppong R, et al.: Optimising outcomes of exercise and corticosteroid injection in patients with subacromial pain (impingement) syndrome: a factorial randomised trial. Br J Sports Med 2021;55:262–71
39. Azadvari M, Emami-Razavi SZ, Torfi F, et al.: Ultrasound-guided versus blind subacromial bursa corticosteroid injection for paraplegic spinal cord injury patients with rotator cuff tendinopathy: a randomized, single-blind clinical trial. Int J Neurosci 2021;131:445–52
40. Kumar PH, Bains TS, Shejale N, et al.: Image-guided versus blind corticosteroid injections in adults with shoulder pain: a systematic review and meta-analysis. Cureus 2021;13:e17032
41. Aly AR, Rajasekaran S, Ashworth N: Ultrasound-guided shoulder girdle injections are more accurate and more effective than landmark-guided injections: a systematic review and meta-analysis. Br J Sports Med 2015;49:1042–9
42. Bal A, Eksioglu E, Gulec B, et al.: Effectiveness of corticosteroid injection in adhesive capsulitis. Clin Rehabil 2008;22:503–12
43. Lee HJ, Lim KB, Kim DY, et al.: Randomized controlled trial for efficacy of intra-articular injection for adhesive capsulitis: ultrasonography-guided versus blind technique. Arch Phys Med Rehabil 2009;90:1997–2002
44. Ekeberg OM, Bautz-Holter E, Tveitå EK, et al.: Subacromial ultrasound guided or systemic steroid injection for rotator cuff disease: randomised double blind study. BMJ 2009;338:a3112
45. Diercks R, Bron C, Dorrestijn O, et al.: Guideline for diagnosis and treatment of subacromial pain syndrome: a multidisciplinary review by the Dutch Orthopaedic Association. Acta Orthop 2014;85:314–22
46. Sharpe RE Jr., Nazarian LN, Levin DC, et al.: The increasing role of nonradiologists in performing ultrasound-guided invasive procedures. J Am Coll Radiol 2013;10:859–63
Keywords:

Subacromial Impingement Syndrome; Adrenal Cortex Hormones; Ultrasonography; Injections; Meta-analysis

Supplemental Digital Content

Copyright © 2021 The Author(s). Published by Wolters Kluwer Health, Inc.