Conservative versus Surgical Interventions for Shoulder Impingement: An Overview of Systematic Reviews of Randomized Controlled Trials

Abstract

Purpose: Numerous systematic reviews (SRs) of randomized controlled trials (RCTs) have emerged that investigate the effectiveness of conservative (supervised exercises) versus surgical (arthroscopic subacromial decompression) interventions for patients with shoulder impingement; however, there are disparities in the quality of the evidence synthesized. The purpose of this study was to conduct an overview of SRs of RCTs to critically appraise the evidence and establish the current state of effectiveness of conservative versus surgical interventions on clinical outcomes among patients with shoulder impingement. Method: The MEDLINE, EMBASE, CINAHL, and PubMed electronic databases were searched for January 2008 to September 2018, and we found SRs of RCTs of patients with shoulder impingement, subacromial pain syndrome, or subacromial impingement syndrome who had received conservative versus surgical interventions to improve outcomes. Two authors extracted the data, and two independent review authors assessed the risk of bias and quality. Results: A total of 15 SRs were identified. One was rated as high quality, 7 as moderate quality, 5 as low quality, and 2 as critically low quality. The results were in line with one another, indicating that no differences in outcomes existed between conservative and surgical interventions among patients with shoulder impingement. Conclusion: There were no clinically important or statistically significant differences in outcomes between conservative versus surgical interventions among patients with subacromial impingement syndrome. To enhance clinical outcomes in this patient population, shoulder-specific exercises that aim to improve muscle strength and flexibility must be considered as the first line of conservative treatment.

Shoulder pain is regarded as the third most prevalent type of musculoskeletal disorder and affects one in three individuals.¹ It has been estimated that the prevalence of rotator cuff–related conditions such as impingement syndrome is 65%–70% in the adult population.^2–3 Shoulder pain may be a major symptom of shoulder impingement syndrome,⁴ and it is considered to be one of the main sources of a reduction in quality of life and a decrease in shoulder joint function.⁵ Studies of the effectiveness of nonsteroidal anti-inflammatory drugs, corticosteroid injection, and conservative approaches (exercise) have been conducted to investigate outcomes in patients with shoulder impingement syndrome.^6–10

Conservative interventions, including exercises, are intended to improve muscle function as well as range of motion by restoring shoulder mobility, proprioception, and stability.¹¹ Numerous systematic reviews (SRs) investigating the effectiveness of conservative versus surgical interventions in patients with shoulder impingement have emerged;^11–25 however, the quality of evidence synthesized by these SRs varies. Given the large increase in the number of published SRs on this topic over the past 3 years, an overview of SRs of randomized controlled trials (RCTs) was warranted to summarize the evidence for use by researchers, clinicians, funding agencies, and policy-makers to assist decision making and evidence translation.

Therefore, we aimed to conduct an overview of SRs of RCTs to critically appraise the evidence and establish the current state of effectiveness of conservative (with an exercise component) versus surgical interventions on clinical outcomes among patients with shoulder impingement.

Methods

Designing the study

This study was an overview of SRs of RCTs. The PROSPERO registration number is CRD42018109357.

Determining the eligibility criteria

We set out to find SRs that met the following population, intervention, comparison, and outcome (PICO) components:

• Population: patients with shoulder impingement, subacromial pain syndrome, or subacromial impingement syndrome (compression of rotator cuff muscles and tendons evident from radiological changes, clinical symptoms, or both)²¹

• Intervention: conservative (with exercise component)

• Comparison: surgical interventions

• Outcome: function or disability, pain, range of motion, and strength.

SRs of non-RCTs; narrative, critical, or scoping reviews; and conference abstracts or posters were excluded.

Conducting the search

An electronic search for SRs published in English between January 2008 and September 2018 was conducted in the following databases: MEDLINE, EMBASE, CINAHL, and PubMed. Moreover, we searched the PROSPERO database and the reference lists of the selected SRs and RCTs to identify other eligible SRs. Our search strategy is summarized in the Appendix.

Selecting the studies

Two independent reviewers (GN and PB) performed the electronic searches in each database. Duplicate SRs were identified and removed. Next, we independently screened the titles and abstracts and retrieved in full text any article marked include or uncertain by either reviewer. Finally, we conducted an independent full-text review to assess final eligibility. In case of disagreement, a third reviewer (JCM) helped achieve consensus through discussion.

Extracting the data

Data extraction was performed by two authors (GN and PB). Descriptive characteristics were extracted from the eligible SRs by including (1) author and year, (2) population, (3) risk of bias, (4) quality of evidence assessment (without rescoring), (5) reported outcomes, and (6) results or conclusions made by the authors of the reviews.

Reporting the quality of the randomized controlled trials in the selected systematic reviews

We did not rescore the quality of the RCTs included in the SRs but instead reported the quality of each SR according to its authors’ assessment. When an RCT was included in two or more SRs, we reported this, along with any variation in the SR authors’ assessments of study quality.

Assessing the quality of the evidence

Several tools were originally used by the SRs selected for our study to assess the quality of the individual studies.

Grading of Recommendations Assessment, Development and Evaluation

The Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach takes into account the risk of bias, publication bias, consistency of findings, precision, and applicability of the overall body of literature to provide a high, moderate, low, or very low rating of the quality of evidence.¹¹

Physiotherapy Evidence Database scale

The Physiotherapy Evidence Database (PEDro) scale uses eight criteria to assess the internal validity of a trial and two criteria to determine the sufficiency of the statistical information displayed in a trial. One criterion is concerned with whether a study provides both point measures and measures of variability. Each criterion can be marked yes or no, where yes = 1 point and no = 0 points on the final rating scale.¹⁶

Scottish Intercollegiate Guidelines Network criteria

The Scottish Intercollegiate Guidelines Network (SIGN) criteria are used to assess the internal validity of an RCT, and assessment involves examining 10 methodological criteria: (1) clarity of the research question, (2) randomization method, (3) appropriate concealment of treatment allocation, (4) blinding of treatment and outcomes, (5) similarity of between-groups baseline characteristics, (6) the co-intervention and contamination involved, (7) the extent of reliability and validity of outcome measures, (8) attrition rate, (9) whether an intention-to-treat analysis was performed, and (10) how comparable the results are across study sites.²⁴

Cochrane Musculoskeletal Group

The Cochrane Musculoskeletal Group (CMSG) guidelines use four levels of quality of evidence: platinum, gold, silver, and bronze. To achieve platinum-level evidence, an SR of at least two RCTs must meet the following criteria: (1) sample size of at least 50 participants per group, (2) patients and outcome assessors blinded, (3) adequate handling of withdrawals of more than 80% at follow-up, and (4) concealment of allocation performed. To achieve gold-level evidence, an SR of at least one RCT must meet the following criteria: (1) sample size of at least 50 participants per group, (2) patients and outcome assessors blinded, (3) adequate handling of withdrawals of more than 80% at follow-up, and (4) concealment of allocation performed. Silver-level evidence is an SR or RCT that does not meet the criteria for platinum- or gold-level evidence. Bronze-level evidence is a high-quality case series without controls or that is derived from expert opinion.²⁵

Assessing the risk of bias

Two independent reviewers (GN and PB) applied the Measurement Tool to Assess Systematic Reviews (AMSTAR 2) risk-of-bias tool to assess the risk of bias in the selected SRs.²⁶ In case of disagreement, a third reviewer (JCM) helped achieve consensus through discussion. AMSTAR 2 scores items in 16 colour-coded domains: inclusion of PICO components, protocol registered before commencement of SR, description of study selection criteria, adequacy of literature search, whether study selection and data extraction were performed in duplicate, justification for exclusion of studies, detailed description of selected studies, assessment of risk of bias for individual studies, statement of source of funding, appropriateness of meta-analytical methods, meta-analysis based on risk of bias, consideration of risk of bias when interpreting the results of the SR, explanation and discussion of heterogeneity, assessment of likely impact of publication bias, and potential sources of conflict of interest.²⁶ Each domain is scored and colour coded as yes (green), partial yes (yellow), no (red), or not applicable (grey).

The overall AMSTAR 2 rating of confidence (quality) can be interpreted as follows:

• High – No, or one, non-critical weakness: the SR provides an accurate and comprehensive summary of the results of the available studies that address the question of interest.²⁶

• Moderate – More than one non-critical weakness: The SR has more than one weakness but no critical flaws. It may provide an accurate summary of the results of the available studies that were included in the review.²⁶

• Low – One critical flaw with or without non-critical weaknesses: The SR has a critical flaw and may not provide an accurate and comprehensive summary of the available studies that address the question of interest.²⁶

• Critically low – More than one critical flaw with or without non-critical weaknesses: The SR has more than one critical flaw and should not be relied on to provide an accurate and comprehensive summary of the available studies.²⁶

Synthesizing the data

We performed a qualitative synthesis of the data and summarized the main results of the selected SRs on the basis of the AMSTAR 2 quality ratings.

Results

Selecting the systematic reviews

Our initial search yielded 455 publications (422 from the database search and 33 from the PROSPERO database and reference lists). After removing duplicates, 281 articles remained and were screened using their title and abstract; this left 43 articles selected for full-text review. Of these, 15 SRs were eligible for our analysis.^11–25 A flowchart showing the selection process is presented in Figure 1.

Figure 1 .

Flowchart of process to select systematic reviews to be selected in the overview.

Characteristics of the selected systematic reviews

The effectiveness of conservative versus surgical interventions on clinical outcomes in patients with shoulder impingement was evaluated in the 15 eligible SRs.^11–25 All 15 SRs reported the outcomes of function or disability and pain,^11–25 7 SRs reported range of motion,^{11–14,16,22,24} and 4 SRs reported strength.^11,14,22,24 The characteristics of the selected SRs are summarized and presented in Table 1.

Table 1 .

Characteristics of Selected Systematic Reviews, Results or Conclusions, and Rating of Confidence Using AMSTAR 2

Author	Population	Studies	Risk-of-bias assessment tool	Quality of evidence* (assessment tool)	Outcomes	Results or conclusions†	AMSTAR 2 quality
Page et al.11	Subacromial impingement syndrome	Brox et al.,27 Haahr et al.28	Cochrane Collaboration’s tool for assessing risk of bias	Very low to low quality (GRADE)	Function, pain, range of motion, global treatment success, adverse events, strength	“All differences in outcomes between supervised exercises and arthroscopic subacromial decompression were not clinically important or statistically significant. Zero events in both groups”11(p.23-24); “Low quality evidence from one trial showed no important differences between exercise (plus heat, cold packs or soft tissue treatment) vs arthroscopic subacromial decompression with respect to overall pain, function, active range of motion and strength at 6- and 12-months, or global treatment success at four to eight years.”11(p.24)	High
Steuri et al.12	Shoulder impingement	Ketola et al.,26 Brox et al.,29 Peters & Kohn,30 Rahme et al.,31 Farfaras et al.,32 Haahr et al.,33 Haahr & Andersen,34 Ketola et al.35	Cochrane Collaboration’s tool for assessing risk of bias	All outcomes – very low quality (GRADE)	Function or disability, pain, range of motion	“Although our review only provides very low-quality evidence, we suggest that exercise may be considered as the core conservative treatment for shoulder impingement. Furthermore, manual therapy, laser and tape might provide additional benefit. Surgery may be a valid alternative after unsuccessful conservative treatments, and for patients with clearly distinguished clinical signs.”12(p.8) “Exercise, especially shoulder-specific exercises, should be prescribed for all patients with shoulder impingement.”12(p.8)	Moderate
Haik et al.16	Subacromial pain syndrome	Brox et al.,27 Haahr et al.28	PEDro scale	All outcomes – moderate quality (GRADE)	Function, pain, range of motion	“According to the body of evidence synthesized, exercise therapy aimed at restoring muscle flexibility and strength of shoulder and scapular muscles should be used as the first-line treatment to improve pain, function and range of motion in individuals with subacromial pain syndrome (SAPS) before recommending arthroscopic surgery.”16(p.11)	Moderate
Goldgrub et al.24	Subacromial impingement syndrome	Haahr et al.,28 Haahr & Andersen34	SIGN criteria	No evidence (best-evidence synthesis)	Function, pain, range of motion, strength	“We did not find evidence (no statistically significant differences) that multimodal care (heat application, cold application, soft tissue therapy, supervised exercises, home exercise) programmes are more effective than surgery followed with exercises for persistent subacromial impingement syndrome.”24(p.136)	Moderate
Saltychev et al.21	Shoulder impingement	Brox et al.,27 Haahr et al.,28 Brox et al.,29 Peters & Kohn,30 Rahme et al.,31 Haahr & Andersen,34 Ketola et al.35	Cochrane Collaboration’s tool for assessing risk of bias	No evidence for all outcomes; pain outcome: moderate quality (best-evidence synthesis)	Function or disability, pain, restriction of movement	“There is no evidence of surgery being more effective than conservative methods in the treatment of shoulder impingement. There is moderate evidence that surgery and conservative methods have similar effect on reduction of pain intensity amongst patients with shoulder impingent in stage two.”24(p.7)	Moderate
Abdulla et al.20	Subacromial impingement syndrome	Ketola et al.35,36	SIGN criteria	Limited or low quality (best-evidence synthesis)	Pain, disability	“For persistent subacromial impingement syndrome, supervised and home-based strengthening exercise leads to similar outcomes as surgery plus post-surgical rehabilitation.”20(p.655)	Moderate
Coghlan et al.25	Impingement, Stage II impingement syndrome	Brox et al.,27 Haahr et al.,28 Rahme et al.31	Cochrane Collaboration’s tool for assessing risk of bias	All outcomes: silver level (CMSG)	Function, pain	“There is evidence from three trials that there is no difference in outcome between surgery and active non-operative treatment for impingement syndrome.”25(p.10)	Moderate
Dorrestijn et al.18	Subacromial impingement syndrome	Brox et al.,27 Haahr et al.,28 Peters & Kohn,30 Rahme et al.,31 Haahr & Andersen34	11-domain list	Medium to low quality (best-evidence synthesis)	Function, pain	“No confident conclusion can be made based on the results available. The randomized controlled trials included in this review failed to provide evidence for differences in outcome between conservatively- and surgically-treated patients with subacromial impingement syndrome (SIS). Whether this failure is due to impairments in methodological quality or a lack of difference in treatment outcome remains unclear.”18(p.53)	Moderate
Dong et al.19	Shoulder impingement syndrome	Haahr et al.28	Cochrane Collaboration’s tool for assessing risk of bias	Not reported	Function, pain	“Exercise and other exercise-based therapies are the most important treatment options for [shoulder impingement syndrome] SIS patients. For those patients who seek nonoperative treatment option[s] at an early stage of SIS, exercise combined with other therapies should be recommended.”19(p.15) “For patients with chronic SIS, operative treatment options may be considered. Notably, however, the decision for operative treatment should be made cautiously because similar outcomes may also be achieved by the implementation of exercise therapy.”19(p.15)	Low
Hanratty et al.15	Subacromial impingement syndrome	Brox et al.,27 Haahr et al.,28 Brox et al.29	Cochrane Collaboration’s tool for assessing risk of bias	Medium quality (best-evidence synthesis)	Function, pain	“No significant differences between physiotherapy vs surgical groups were found. Both exercise and surgery were significantly better than placebo in reducing pain and improving function; neither treatment was superior.”15(p.6-8)	Low
Kelly et al.22	Subacromial impingement syndrome	Brox et al.,27 Haahr et al.,28 Brox et al.,29 Rahme et al.31	Unclear	Overall poor (PEDro scale)	Function or disability, pain, strength, range of motion	“No significant differences between physiotherapy and surgical groups.”22(p.106) “The review has shown exercise to be effective to some degree in the management of subacromial impingement syndrome. Support for exercise can only be tentatively accepted because of methodological flaws in the included studies.”22(p.107)	Low
Braun et al.23	Shoulder impingement	Ketola et al.35	Unclear	PEDro scale	Pain, disability	“There is limited evidence from one RCT to indicate that arthroscopic subacromial decompression surgery may not provide superior effects compared with a physiotherapeutic exercise regime.”23(p.282)	Low
Gebremariam et al.17	Subacromial impingement syndrome	Brox et al.,27 Haahr et al.,28 Rahme et al.31	Furlan’s 12 quality criteria	Low quality (Furlan quality criteria)	Function, pain	“No evidence was found for the superiority of subacromial decompression (arthroscopic or open) compared with conservative treatment in the short, mid, and long term.”17(p.1905)	Low
Kuhn13	Impingement syndrome	Brox et al.,27 Haahr et al.,28 Brox et al.,29 Peters & Kohn,30 Rahme et al.31	3-domain list (randomization, independent examiner, and follow-up assessed)	Unclear; Levels 1 and 2	Function or disability, pain, range of motion	“All studies failed to show statistically significant differences between the acromioplasty with exercise vs exercise alone treatments.”13(p.148)	Critically low
Kromer et al.14	Subacromial impingement syndrome	Brox et al.27 Haahr et al.,28 Brox et al.,29 Haahr & Andersen34	PEDro scale	Moderate to strong (best-evidence synthesis)	Function or disability, pain, range of motion, strength	“According to our best-evidence synthesis, moderate evidence was found for an equal effectiveness of physiotherapist-led exercises and surgery in patients with subacromial impingement syndrome (SIS), especially in the long term.”14(p.878)	Critically low

^*Quality of evidence per outcome for individual studies included in each systematic review, according to its authors.

^†According to the authors of the reviews.

AMSTAR 2 = Measurement Tool to Assess Systematic Reviews; GRADE = Grading of Recommendations Assessment, Development and Evaluation; PEDro = Physiotherapy Evidence Database; SIGN = Scottish Intercollegiate Guidelines Network; CMSG = Cochrane Musculoskeletal Group.

Risk-of-bias assessment in the selected systematic reviews

The selected SRs evaluated the risk of bias using the following assessment tools:

• CMSG guidelines: six SRs^{11–12,15,19,21,25}

• PEDro scale: two SRs^14,16

• SIGN: two SRs^20,24

• Furlan: one SR¹⁷

• 11-domain list: one SR¹⁸

• 3-domain list: one SR¹³

• Unclear (not reported): two SRs.^22–23

Assessing the quality of evidence in the selected systematic reviews

The quality of evidence was assessed using the following assessment tools:

• GRADE: three SRs^11–12,16

• Best-evidence synthesis approach: six SRs^{14–15,18,20–21,24}

• PEDro scale: two SRs^22–23

• CMSG guidelines: one SR²⁵

• Furlan criteria: one SR¹⁷

• Not reported: one SR¹⁹

• Unclear: one SR¹³

Assessing the quality of the systematic reviews using the Measurement Tool to Assess Systematic Reviews 2

Using AMSTAR 2, we rated the quality of the 15 SRs as follows: 1, high;¹¹ 7, moderate;^{12,16,18,20,21,24,25} 5, low;^{15,17,19,22,23} and 2, critically low.^13–14 Regarding the tool’s critical domains, 10 SRs did not perform a priori registration,^{12–14,16–19,21–23} and 2 partially met the comprehensive search criteria.^16,17 Two SRs did not provide justification for the excluded studies,^13,21 and 7 partially used a satisfactory technique for assessing the risk of bias of the primary studies.^{14,16,18,20,22–24} Six SRs did not consider the risk of bias when interpreting their results,^{13,14,17,19,22,23} and 4 did not assess the presence or likely impact of publication bias in the selected studies.^14,15,19,25 A summary of the AMSTAR 2 ratings is presented in Table 2.

Table 2 .

AMSTAR 2 Ratings for the 15 Selected Systematic Reviews

Domain	Page et al.11	Steuri et al.12	Haik et al.16	Goldgrub et al.24	Saltychev et al.21	Abdulla et al.20	Coghlan et al.25	Dorrestijn et al.18	Dong et al.19	Hanratty et al.15	Kelly et al.22	Braun et al.23	Gebremariam et al.17	Kuhn13	Kromer et al.14
1. Research question and inclusion criteria aligned with PICO	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
2. A priori protocol	Yes	No	No	Partial yes	No	Yes	Yes	No	No	Yes	No	No	No	No	No
3. Study design selection explained	No	No	No	No	No	No	No	Yes	No	No	No	No	No	No	Yes
4. Comprehensive search	Yes	Yes	Partial yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Partial yes	Yes	Yes
5. Duplicate study selection	Yes	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	No	Yes	Yes	Yes	No	No
6. Duplicate data extraction	Yes	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	No
7. List and justification of excluded studies	Yes	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	No	Yes
8. Included studies described in adequate detail	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Partial yes	Partial yes	Yes	Yes	Yes	Partial yes	Yes	Yes
9. Satisfactory technique for assessment of risk of bias	Yes	Yes	Partial yes	Partial yes	Yes	Partial yes	Yes	Partial yes	Yes	Yes	Partial yes	Partial yes	Yes	No	Partial yes
10. Sources of funding of included studies reported in review	Yes	No	No	No	No	No	Yes	No	No	No	No	No	No	No	No
11. If meta-analysis: justified combination of data	Yes	Yes	Not applicable	Not applicable	Yes	Not applicable	Yes	Not applicable	Yes	No	Not applicable	Not applicable	Not applicable	Not applicable	No
12. If meta-analysis: risk of bias of included studies taken into account	Yes	Yes	Not applicable	Not applicable	Yes	Not applicable	Yes	Not applicable	Yes	No	Not applicable	Not applicable	Not applicable	Not applicable	No
13. Risk of bias taken into account in interpretation and discussion	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	No	Yes	No	No	No	No	No
14. Satisfactory explanation for any heterogeneity	Yes	Yes	No	No	Yes	No	Yes	No	Yes	No	No	No	No	No	No
15. Publication bias in included studies assessed	Yes	Yes	Not applicable	Not applicable	Yes	Not applicable	No	Not applicable	No	No	Not applicable	Not applicable	Not applicable	Not applicable	No
16. Review authors report on any of their own conflicts of interest	Yes	Yes	Yes	Yes	Yes	No	Yes	No	Yes	Yes	No	Yes	Yes	No	No
Overall quality rating	High	Moderate	Moderate	Moderate	Moderate	Moderate	Moderate	Moderate	Low	Low	Low	Low	Low	Critically low	Critically low

Note: Each of the 16 criteria on the AMSTAR 2 checklist are shown for each of the 15 selected systematic reviews. Yes, no, partial yes, and not applicable indicate whether the study satisfied each of these criteria.

AMSTAR 2 = Measurement Tool to Assess Systematic Reviews; PICO = population, intervention, comparison, and outcome.

High-quality systematic reviews

We rated one SR as high-quality evidence.¹¹ This SR included very low- to low-quality RCTs and used the Cochrane Collaboration’s tool for assessing the risk of bias and the GRADE approach for assessing the quality of evidence. It showed no clinically important or statistically significant differences in outcomes (function, pain, range of motion, strength) between supervised exercises and arthroscopic subacromial decompression in patients with subacromial impingement syndrome at 6 and 12 months or global treatment success at 4–8 years.¹¹ Moreover, no events were reported in either group.

Moderate-quality SRs

We rated seven SRs as having moderate-quality evidence.^{12,16,18,20,21,24,25} Using GRADE, one SR included very low-quality RCTs.¹² It showed no differences in clinical outcomes (function, pain, range of motion) between exercise and surgery among patients with shoulder impingement syndrome at the shortest and longest follow-ups.¹² In addition, one SR included moderate-quality RCTs and showed no differences in outcomes (function, pain, range of motion) between exercise and arthroscopic surgery in patients with subacromial pain syndrome at 6 and 12 months.¹⁶

Using the best-evidence synthesis approach, four SRs included limited- to moderate-quality RCTs and indicated no differences in clinical outcomes (function, pain, range of motion, strength) between exercise and surgery in patients with subacromial impingement syndrome at 6 months to 4 years.^18,20,21,24 Using the CMSG guidelines, one SR included silver-level evidence RCTs and showed no differences in outcomes (function, pain) between exercise and surgery in patients with impingement syndrome at 6 and 12 months.²⁵

Low-quality SRs

We rated five SRs as having low-quality evidence.^{15,17,19,22,23} Using the best-evidence synthesis approach, one SR included moderate-quality RCTs and indicated no differences in outcomes (function, pain) between exercise and arthroscopic surgery in patients with subacromial impingement syndrome at 6 and 12 months.¹⁵ Using the Furlan quality criteria,¹⁷ one SR included low-quality RCTs and determined there were no differences in outcomes (function, pain) between exercise and arthroscopic surgery among patients with subacromial impingement syndrome at short-, mid-, and long-term follow-up.¹⁷

Using the PEDro scale, one SR of poor-quality RCTs found no differences in outcomes (function, pain, range of motion, and strength) between physiotherapy (exercise) and surgery in patients with subacromial impingement syndrome at 6 and 12 months.²² Two SRs, each with one RCT, showed no differences in outcomes (function, pain) between exercise and surgery among patients with shoulder impingement syndrome at 6 and 12 months.^19,23 The quality of the individual RCTs in these SRs was not assessed.^19,23

Critically low-quality SRs

We rated two SRs as having critically low-quality evidence.^13,14 Using a three-domain list, one SR included Levels 1 and 2 quality RCTs; it indicated no differences in outcomes (function, pain, range of motion) between exercise and surgery among patients with impingement syndrome at up to 4 years after intervention.¹³ Using the best-evidence synthesis approach, one SR included moderate- to strong-quality RCTs and indicated no differences in outcomes (function, pain, range of motion, and strength) between exercise and surgery among patients with subacromial impingement syndrome for up to 4 years.¹⁴

Discussion

We conducted an overview of SRs of RCTs to critically appraise the evidence and establish the current state of the effectiveness of conservative versus surgical interventions on clinical outcomes among patients with shoulder impingement. Our overview indicates that the results from the 15 selected SRs were in line with one another – stating that there were no differences in clinical outcomes between supervised exercises and surgery among patients with shoulder impingement. However, we noted disparities in their methodological quality. Such disparities influence the extent to which an accurate and comprehensive summary of the results can be provided.

The SRs used various appraisal tools to assess the risk of bias in the included trials and to summarize the quality of evidence. We were able to demonstrate that variations in the appraisal tools used affected the quality of evidence synthesized from the SRs. Haik and colleagues used the PEDro scale and the GRADE approach to synthesize moderate-quality evidence;¹⁶ Kromer and colleagues used the PEDro scale and best-evidence synthesis approach and reported moderate- to strong-quality evidence.¹⁴ Both SRs were based on the same RCTs across the same outcomes. This highlights the variations in the quality of evidence synthesized depending on the type of appraisal tool used.

To synthesize the quality of evidence pertaining to effectiveness, trials must use the Cochrane Collaboration’s tool for assessing risk of bias, along with the GRADE approach; using other quality scales along with the GRADE approach is not appropriate.¹¹ Moreover, we noted a lack of reporting and distinction between assessments of risk of bias and assessments of quality. The terms were often used interchangeably in the selected SRs. Quality refers to the degree to which researchers conduct their study to the highest possible standards, whereas risk-of-bias assessment pertains to the extent to which the results of a study should be believed. In addition, we identified 10 SRs that had appraised individual clinical trials using quality scales with resulting scores.^{13–18,20,22–24} This method has been deemed inappropriate because it tends to combine aspects of the quality of reporting with aspects of conducting trials. Page and colleagues’ high-quality SR and Steuri and colleagues’ moderate-quality SR used identical appraisal tools to assess quality and risk of bias but were rated differently using AMSTAR 2.^11,12 This indicates that the methodological quality used in each SR was not merely based on the choice of appraisal tools used.

Well-conducted SRs provide the highest level of evidence from which strong inference can be drawn.^37–41 In this overview, we identified 13 SRs that did not report on the sources of funding for the trials they included. There is evidence to support the hypothesis that the results of industry-funded trials tend to favour sponsored products and that such trials are less likely to be published than trials that are independently funded.²⁶ Moreover, we identified 10 SRs without protocol registrations. To reduce the risk of bias, it is important that review authors develop, agree on, and register the methods (i.e., protocol registration) of their SR before they carry it out.²⁶

There have been no other overviews on the effectiveness of conservative versus surgical interventions in patients with shoulder impingement. Therefore, we were unable to compare our findings.

Among patients with shoulder or subacromial impingement syndrome, shoulder-specific exercises that aim to enhance muscle strength and flexibility must be considered the first line of conservative treatment to improve clinical outcomes. Large-scale, multi-centre, well-designed RCTs are required to further support these findings.

This overview had one limitation: it did not include observational studies, conference papers, posters, or abstracts. This approach potentially contributed to a publication bias in our search strategy; however, our aim was to summarize the highest level of evidence available.

Conclusions

This overview of SRs pooled 15 reviews, ranging from critically low to high quality, and found that there were no clinically important or statistically significant differences in outcomes between conservative (supervised exercises) versus surgical interventions (arthroscopic subacromial decompression) among patients with subacromial impingement syndrome.

Key Messages

What is already known on this topic

Numerous systematic reviews (SRs) have been published on the effectiveness of conservative versus surgical interventions in patients with shoulder impingement, but the quality of the evidence synthesized by these SRs varies.

What this study adds

This overview of SRs of randomized controlled trials summarized, critically appraised, and established the current state of effectiveness of conservative (with exercise component) versus surgical interventions on clinical outcomes in patients with shoulder impingement. We found that shoulder-specific exercises must be considered the first line of conservative treatment to improve clinical outcomes.

Appendix: Search Strategy

Database	Search condition	No. of results
MEDLINE	1 shoulder impingement	79
	2 shoulder impingement syndrome
	3 shoulder pain
	4 subacromial pain syndrome
	5 supraspinatus
	6 painful arc syndrome
	7 painful arc
	8 subacromial pain syndrome
	9 1 OR 2 OR 3 OR 4 OR 5 OR 6 OR 7 OR 8
	10 shoulder surgery
	11 arthroscopic
	12 subacromial
	13 decompression
	14 arthroscopic subacromial decompression
	15 open subacromial decompression
	16 acromioplasty
	17 shoulder repair
	18 debride
	19 shoulder surgical
	20 10 OR 11 OR 12 OR 13 OR 14 OR 15 OR 16 OR 17 OR 18 OR 19
	21 conservative
	22 exercise
	23 physiotherapy
	24 physical therapy
	25 therapy
	26 rehabilitation
	27 strengthening exercises
	28 stretching exercises
	29 21 OR 22 OR 23 OR 24 OR 25 OR 26 OR 27 OR 28
	30 systematic review
	31 9 AND 20 AND 29 AND 30
EMBASE	1 shoulder impingement	47
	2 shoulder impingement syndrome
	3 shoulder pain
	4 subacromial pain syndrome
	5 supraspinatus
	6 painful arc syndrome
	7 painful arc
	8 subacromial pain syndrome
	9 1 OR 2 OR 3 OR 4 OR 5 OR 6 OR 7 OR 8
	10 shoulder surgery
	11 arthroscopic
	12 subacromial
	13 decompression
	14 arthroscopic subacromial decompression
	15 open subacromial decompression
	16 acromioplasty
	17 shoulder repair
	18 debride
	19 shoulder surgical
	20 10 OR 11 OR 12 OR 13 OR 14 OR 15 OR 16 OR 17 OR 18 OR 19
	21 conservative
	22 exercise
	23 physiotherapy
	24 physical therapy
	25 therapy
	26 rehabilitation
	27 strengthening exercises
	28 stretching exercises
	29 21 OR 22 OR 23 OR 24 OR 25 OR 26 OR 27 OR 28
	30 systematic review
	31 9 AND 20 AND 29 AND 30
CINAHL	1 shoulder impingement	31
	2 shoulder impingement syndrome
	3 shoulder pain
	4 subacromial pain syndrome
	5 supraspinatus
	6 painful arc syndrome
	7 painful arc
	8 subacromial pain syndrome
	9 1 OR 2 OR 3 OR 4 OR 5 OR 6 OR 7 OR 8
	10 shoulder surgery
	11 arthroscopic
	12 subacromial
	13 decompression
	14 arthroscopic subacromial decompression
	15 open subacromial decompression
	16 acromioplasty
	17 shoulder repair
	18 debride
	19 shoulder surgical
	20 10 OR 11 OR 12 OR 13 OR 14 OR 15 OR 16 OR 17 OR 18 OR 19
	21 conservative
	22 exercise
	23 physiotherapy
	24 physical therapy
	25 therapy
	26 rehabilitation
	27 strengthening exercises
	28 stretching exercises
	29 21 OR 22 OR 23 OR 24 OR 25 OR 26 OR 27 OR 28
	30 systematic review
	31 9 AND 20 AND 29 AND 30
PubMed	1 shoulder impingement	265
	2 shoulder impingement syndrome
	3 shoulder pain
	4 subacromial pain syndrome
	5 supraspinatus
	6 painful arc syndrome
	7 painful arc
	8 subacromial pain syndrome
	9 1 OR 2 OR 3 OR 4 OR 5 OR 6 OR 7 OR 8
	10 shoulder surgery
	11 arthroscopic
	12 subacromial
	13 decompression
	14 arthroscopic subacromial decompression
	15 open subacromial decompression
	16 acromioplasty
	17 shoulder repair
	18 debride
	19 shoulder surgical
	20 10 OR 11 OR 12 OR 13 OR 14 OR 15 OR 16 OR 17 OR 18 OR 19
	21 conservative
	22 exercise
	23 physiotherapy
	24 physical therapy
	25 therapy
	26 rehabilitation
	27 strengthening exercises
	28 stretching exercises
	29 21 OR 22 OR 23 OR 24 OR 25 OR 26 OR 27 OR 28
	30 systematic review
	31 9 AND 20 AND 29 AND 30