Reporting and Methodological Quality of Systematic Reviews and Meta-Analyses Evaluating Effects of Extracorporeal Shock Wave Therapy on Tendinopathies: A Scoping Review

Abstract

Objective

The purpose of this study was to evaluate the methodological and reporting quality of the published systematic reviews with or without meta-analyses (SRs/MAs) that looked at the effects of extracorporeal shock wave therapy (ESWT) on tendinopathies and to summarize its effectiveness.

Methods

A search of PubMed, Scopus, PEDro, Web of Science, Cochrane, Embase, and REHABDATA was conducted. SRs/MAs that assessed the effectiveness of ESWT for treating tendinopathy were included. The methodological and reporting quality of the eligible SRs/MAs were assessed using AMSTAR-2 and the PRISMA checklist. In addition, the ROBIS tool was applied to evaluate the risk of bias (RoB).

Results

Eighteen SRs/MAs were included. The overall methodological quality was “critically low.” Furthermore, the reporting quality of the included reviews according to PRISMA criteria was not optimal. Based on the ROBIS, a total of 16.2% of the studies had a low RoB, 38.9% had an unclear RoB, and 44.4% of the studies were appraised as having a high RoB.

Conclusion

In this scoping review we found substantial limitations regarding the quality and RoB of SRs/MAs. Therefore, reviewers must consider the AMSTAR-2, PRISMA, and ROBIS tools to improve the quality of future studies.

Introduction

Tendinopathies of the upper and lower limbs are painful and reduce the range of motion. They burden society because they result in healthcare costs.¹ Tendinopathy is characterized by increased tendon thickness, disorganized fibrils, and neovascularization caused by repetitive microtrauma, which is most commonly caused by loading activities such as exercise.² As a result, developing an effective intervention is essential.

Tendinopathy is treated with a wide range of therapies, including conservative therapies such as physiotherapy and nonsteroidal anti-inflammatory drugs.^3,4 If initial conservative treatment fails, other treatments such as corticosteroid injections or surgery may be prescribed.^5,6 The extracorporeal shock wave therapy (ESWT) technique has been suggested as an alternative treatment for tendinopathy management to conservative treatments or surgery.^7,8 ESWT has been widely used in clinical practice for more than 3 decades to manage musculoskeletal disorders, most of which are tendinopathies.^{9, 10, 11} This modality uses pulsed acoustic waves with high or low energy to induce rapid pressure changes in tissues. This stimulus starts a chain of biochemical reactions in the targeted tissues, which helps the tissues heal and grow back.¹²

There have been many studies on how well ESWT works to treat tendinopathy, and these results have been summed up in a number of systematic reviews and meta-analyses (SRs/MAs).^{13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30} Evidence from SRs is considered the gold standard for assessing the effectiveness of an intervention.³¹ Since clinical decisions are reliant on information from SRs with or without MAs, using limited or incorrect data might be harmful for patients. It is vital to critically assess the quality of SRs of interest as part of an evidence-based approach in order to validate their accuracy and dependability.

Several checklists and tools have been developed to provide some instructions to authors in order to enhance the reporting quality of review studies, such as “Preferred Reporting Items for SRs/MAs (PRISMA)”.³² Additionally, a number of tools like “A Measurement Tool to Assess SRs 2 (AMSTAR-2)”³³ have been established to strengthen the methodological quality of SRs. Both tools make it possible to find flaws in SR studies that are already done and take steps to improve reporting and methodology quality in future reviews.

The purpose of this scoping review was to evaluate the methodological and reporting quality of the published SRs/MAs that looked at the effectiveness of ESWT for treating tendinopathy, and the secondary aim was to summarize its efficacy.

Methods

The PRISMA Extension for Scoping Reviews (PRISMA-ScR) checklist was considered in designing, conducting, and reporting this review study.³⁴

Search Strategy

The search strings were created employing 3 components of PICO-S (population, intervention, comparison, outcome, and study design), including population (subjects with tendinopathy), intervention (ESWT), and study design (SRs/MAs). Firstly, Medical Subject Headings and Emtree thesaurus were looked for to identify relevant terms. Then, similar research looked at common terms and contacted professionals in the field to seek other potential terms. The search string was generated for the PubMed database first and subsequently adapted for use on the other electronic databases (See Supplementary file Table 1).

From inception to January 30, 2022, PubMed, Scopus, Web of Science, and Embase were searched. Furthermore, PEDro (the Physiotherapy Evidence Database), the Cochrane Database of Systematic Reviews, and REHABDATA were searched manually. Various databases, such as Microsoft Academic, OpenGrey, and Google Scholar, were explored to find relevant gray literature. During the search, language was not a factor. If a study was found in a language other than English, the abstract was translated using Google Translate. Also, the reference lists of the studies that were included and relevant major journals (like Archives of Physical Medicine and Rehabilitation, Clinical Rehabilitation, BMC Musculoskeletal Disorders, The American Journal of Sports Medicine, etc.) were searched for any possible research. The search process was performed by the first author (S.Sh.).

After searching the electronic databases, the first records were put into the Endnote X8 software (Thomson Reuters, New York, NY). The duplicates were then deleted, and the remaining studies were screened by 2 authors (S. Sh. and Sh. Sh.) independently based on the title and abstract. Following that, the same 2 authors examined the entire text of probable related studies based on the inclusion and exclusion criteria, and the final papers were chosen for inclusion in the study. If there were any disagreements during this process, they would be discussed, and if necessary, a third author (K.B.L.) would be brought in.

Eligibility Criteria

The research team considered SRs/MAs evaluating the effects of ESWT on tendinopathies. The intended outcomes included pain relief, function, disappearance of calcifications, satisfaction, and quality of life (QoL). In the following comparisons, the eligible studies had at least examined the effects of ESWT on tendinopathies as one of the key primary goals: (1) ESWT versus no intervention, (2) ESWT versus placebo, (3) ESWT versus surgical or non-surgical interventions, (4) ESWT versus other types of shock wave therapy, and (5) ESWT versus an active treatment such as injections. On the other hand, conference abstracts, in vitro studies, SRs/MAs of animal studies, and protocol studies were excluded.

Data Extraction

The authors (S. Sh. and Sh. Sh.) extracted the data independently. The following information was taken from the studies that were included: (1) first author; (2) publication year; (3) journal; (4) type of SR; (5) population; (6) intervention arm; (7) comparison arm; (8) outcomes; (9) number of included studies; (10) number of study participants; (11) risk-of-bias assessment tool; (12) funding source; and (13) summary of findings. Any disagreements between the authors were handled through discussion and the participation of the third author (K. BL), as in the previous steps.

Evaluation of Methodological and Reporting Quality

The AMSTAR-2 tool³⁵ and the PRISMA checklist (version 2020)³⁶ were used to assess the methodological and reporting quality of the included reviews. AMSTAR-2 can be used to evaluate the methodological quality of SRs from both randomized and non-randomized controlled trials.³⁵ This tool has 16 items, and the overall ranking is based on the deficits in 7 crucial domains, which include items 2, 4, 7, 9, 11, 13, and 15. Based on the AMSTAR-2 tool, we ranked the overall methodological quality of the included studies as follows: (1) high, with no defects in any crucial domains; however, non-crucial domains could have up to 3 defects; (2) moderate, with more than 3 defects in non-crucial domains; (3) low, with 1 significant defect in a crucial domain; and (4) critically low, with more than 1 significant defect in crucial domains. The reporting quality of included SRs was assessed using the PRISMA checklist (version 2020).³⁶ There are 27 items in this reporting guidance, covering seven sections (title, abstract, introduction, methods, results, discussion, and other information). Each item was judged as “yes” (complete reports), “partial yes” (partial reports), or “no” (no reports).

Evaluation of Risk of Bias

The ROBIS tool, which assesses the risk of bias (RoB) in all types of SRs, was recently published and validated.³⁷ This tool has 3 parts: (1) assessing relevance (optional), (2) identifying concerns with the review process, and (3) judging the RoB. Since the first section is optional, the research team did not take this into account in the current study. Using the ROBIS tool, the total amount of bias concerns about each domain found by the signaling question were put into 1 of 3 categories: “high,” “unclear,” or “low.” Two authors (S. Sh. and Sh. Sh.) independently evaluated the quality and RoB of included studies. Any discrepancies were resolved through discussion and the involvement of the third author (K. B. L.).

Forest plots were used to show that the included studies met the requirements of the AMSTAR-2 tool and the PRISMA checklist. Two-sided 95% Confidence Intervals (CIs) of proportions were calculated employing the Wilson Score method. Statistical analysis was also conducted using Stata 15.0 software (StataCorp LP, College Station).³⁸

Results

The initial search of the electronic databases turned up 356 studies, and after the duplicates were taken out, only 141 were looked at. Following the screening, 33 studies were retained. The full texts of these studies were looked at, and 18 of them were finally accepted into the study (Fig 1).^{13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30} The excluded studies are listed in Supplementary Table 2 along with the reason why they were excluded.

Fig 1.

PRISMA Flow chart. The PRISMA flowchart demonstrating the searching, screening, and selecting the final studied.

Characteristics of Included Reviews

The characteristics of the final included SRs/MAs are summarized in Table 1. Eight of the 18 studies that were included were SRs;^{13,14,17, 18, 19,22,25,28} seven were SRs/Mas;^{15,20,21,23,24,26,27} and 3 were only meta-analyses.^16,29,30 Pain and function were the most frequently examined outcomes in the included studies (Table 1). Additionally, many of the included studies applied the Cochrane and PEDro risk of bias assessment tools for evaluating the RoB status. Some reviews, however, used other relevant tools like the Newcastle-Ottawa Scale, the Oxford tool, the Modified Downs and Black Checklist, and the Jadad scale.

Table 1.

Characteristics of Included Reviews

First author (year)	Country of Corresponding Author	Journal	Type	Population	Intervention	Comparison	Outcome(s)	Number of Studies Included	Number of Study Participants	Risk-of-bias assessment Tool	Summary of Findings of Review	Funding Source
Al-Abbad et al. (2013) 21	Saudi Arabia	Foot and Ankle International	SR of RCTs and pre-post study designs	Chronic AT	ESWT	Placebo, or other nonsurgical intervention	Pain, function	6	365 (treated tendons)	PEDro Risk of bias assessment tool	ESWT was effective in reducing pain and increasing function at a minimum 3 months’ follow-up	No funding.
Bannuru et al. (2014) 22	USA	Annals of Internet Medicine	SR of RCTs	Shoulder tendinopathies	High-energy ESWT	Low-energy ESWT or placebo	Pain, function, resolution of calcifications	31	1745	Cochrane Risk of bias assessment tool	High-energy ESWT was significantly better than low-energy ESWT and placebo in treatment of calcific tendinitis. ESWT was not effective in treatment of non-calcific tendinitis, regardless of energy dose.	No funding.
Buchbinder et al. (2006) 23	Australia	The Journal of Rheumatology	SR and MA of RCTs	Tennis elbow	ESWT	placebo, or another modality	Pain, function/disability, satisfaction with treatment, QOL	10	1099	Cochrane Risk of bias assessment tool	ESWT was no more effective than placebo or other modalities for tennis elbow treatment	NI
Fan et al. (2020) 24	China	Orthopaedic journal of sports medicine	SR and MA of RCTs and case-control studies	AT	ESWT	Placebo, or other nonsurgical intervention	Pain, function	5 RCT, 3 case-control	442	Cochrane Risk of bias assessment tool, Newcastle-Ottawa Scale	ESWT was more effective than other treatments at both low-energy level and medium-energy level and at both shorter (<6 months) and longer (6 months) follow-up	Multiple funding.
Fatima et al. (2021) 25	Pakistan	The Journal of Pakistan Medical Association	SR and MAof RCTs	Rotator cuff tendinopathy	ESWT	Placebo, or traditional rehabilitation	Pain, function	11	531	Cochrane Risk of bias assessment tool	ESWT was no more effective than traditional rehabilitation for rotator cuff tendinopathy treatment	No funding.
Harniman et al. (2004) 26	Canada	Journal of Hand Therapy	SR of all study types	Rotator cuff tendinopathy	High- energy ESWT	Low- energy ESWT, placebo	Pain	16	1228	Oxford tool	High energy ESWT was effective in treating calcific rotator cuff tendonitis. Low energy ESWT was not effective for treating non-calcific rotator cuff tendonitis	NI
Huisstede et al. (2011) 27	Netherlands	Manual Therapy	SR of RCTs	Rotator cuff tendinopathy	ESWT	ESWT (Low- High or medium energy), no treatment, placebo, or other nonsurgical intervention	Pain, function, or recovery	17	1075	Cochrane Risk of bias assessment tool	High-energy ESWT was significantly more effective than low-energy ESWT (in short-term) and placebo (in short-, mid- and long-term) in treating calcific rotator cuff tendonitis. ESWT was not effective in treatment of non-calcific tendinitis, regardless of energy dose.	NI
Ioppolo et al. (2013) 28	Italy	Archives of Physical Medicine and Rehabilitation	SR and MA of RCTs	Calcific tendinitis of the shoulder	ESWT	Placebo or no treatment	Pain, function, disappearance of calcifications	6	460	PEDro Risk of bias assessment tool	High-energy ESWT was effective in reducing pain, increasing shoulder function, and dissolving calcifications at 6 months’ follow-up	No funding.
Lafrance et al. (2019) 29	Canada	BMJ Open Sport and Exercise Medicine	SR and MA of RCTs	Rotator cuff calcific tendinopathy	ultrasound-guided lavage alone or in conjunction with another intervention	ESWT, RPT, corticosteroid injection	Pain, function	3	274	Cochrane Risk of bias assessment tool	Ultrasound-guided lavage was more effective than ESWT in reducing pain and improving function.	Institut de recherche Robert-Sauvé en santé et sécurité au travail (IRSST)
Lee et al. (2011) 30	Australia	Journal of Shoulder and Elbow Surgery	SR of RCTs	Chronic calcific shoulder tendinitis	ESWT	ESWT (Low or High energy), placebo,	Pain, function	9	589	PEDro Risk of bias assessment tool	ESWT was effective in reducing pain and increasing shoulder function for up to a year after its application	No funding.
Liao et al. (2018a) 31	Taiwan	BMC Musculoskeletal Disorders	SR and MA of RCTs	knee tendinopathies and other soft tissue disorders	ESWT	Placebo, or other nonsurgical intervention	Pain, ROM, treatment success rate	19	1189	PEDro Risk of bias assessment tool	ESWT was effective in reducing pain and ROM restoration. Low energy F-ESWT had greater efficacy for the treatment success rate than high energy F-ESWT, whereas the inverse result was observed for RPT	No funding.
Liao et al. (2018b) 32	Taiwan	American Journal of Physical Medicine and Rehabilitation	SR and MA of RCTs	Lower-Limb tendinopathies	ESWT	Placebo or active treatment (e.g., exercise, injections, or surgery)	Pain, treatment success rate, function/disability	29	1865	PEDro Risk of bias assessment tool	ESWT was effective in reducing pain and increasing lower-limb function. High energy F-ESWT, high- energy RPT, and low-dosage RPT had superior pooled effects on overall clinical outcomes	NI
Louwerens et al. (2016) 33	Netherlands	Arthroscopy	SR of RCTs, non-RCTs, cohort study designs	Chronic calcific rotator cuff tendinopathy	High energy ESWT, Ultrasound-guided needling,	arthroscopic surgery	Pain, function, disappearance of calcifications	22	1,258 (treated shoulders)	Coleman scoring system	High-energy ESWT, US-guided needling, or arthroscopic surgery, were effective in improving clinical outcomes after 1-year follow-up	NI
Mani-Babu et al. (2015) 34	United Kingdom	The American Journal of Sports Medicine	SR and MA of all study designs	Lower-Limb tendinopathies	ESWT	Without control, placebo, any treatment	Pain, function	20	NI	Modified Downs and Black checklist	ESWT was more effective than non-operative treatments for lower-limb tendinopathies and equal to patellar tenotomy surgery for patellar tendinopathy	No funding.
Punnoose et al. (2017) 35	United Kingdom	Journal of Physiotherapy and Physical Rehabilitation	SR and MA of RCTs	AT and Patellar Tendinopathy	ESWT	Placebo, Eccentric exercises	Pain, function	9	487	PEDro Risk of bias assessment tool	ESWT was no more effective than placebo or other treatment.	No funding.
Saithna et al. (2009) 36	United Kingdom	Current Orthopaedic Practice	SR of RCTs	Calcific rotator cuff tendinopathy	ESWT	Placebo	Pain, function	3	260	CONSORT checklist	ESWT was more effective than placebo in improving clinical outcomes at 6 months	NI
Vavken et al. (2009) 37	Austria	Sports Health	SR and MA of controlled trials	Calcific shoulder tendinitis	ESWT	Placebo, any treatment	Pain, function, disappearance of calcifications	14	995	Jadad scale	ESWT was more effective than placebo or other treatments in improving clinical outcomes. High-energy ESWT produced significantly better results than low-energy ESWT for pain reduction and improvement of function	NI
Verstraelen et al. (2014) 38	Netherlands	Clinical Orthopaedics and Related Research	SR and MA of RCTs	Calcific shoulder tendinitis	High energy ESWT	low energy ESWT	Pain, function, disappearance of calcifications	5	410	Cochrane Risk of bias assessment tool	High energy ESWT was more effective than low energy ESWT in improving clinical outcomes at 3 and 6 months	No funding.

At, Achilles tendinopathy; ESWT, extracorporeal shock wave therapy; F-ESWT, focused extracorporeal shockwave therapy; MA, meta-analysis; NI, no information; PEDro, physiotherapy evidence database scale; PHT, proximal hamstring tendinopathy; PRP, platelet-rich plasma; PT, patellar tendinopathy; QOL, quality of life2; RCT, randomized control trial; ROBINS-I, risk of bias in non-randomized studies; ROM, range of motion; RPT, radial pulse therapy; SR, Systematic review; VISA-P, Victorian Institute of Sports Assessment-Patella.

Results of Methodological Quality Evaluation

Table 2 shows the findings of each SR's methodological quality as evaluated by the ASMTAR-2 items. Notably, the overall methodological quality of all included reviews was critically low. Of all the items, only items 14, 5, 6, 13, 11, 16, and 1 were fully adhered to by more than half of the studies that were included (Figure 2). To be more specific, the second item of the AMSTAR-2 tool, which involves posting the pre-study protocol and any significant changes from it, had the lowest level of compliance. None of the studies that were looked at mentioned this item as a key domain. In addition, items 4, 7, and 9, which were also other critical domains, were not well covered by the included studies. Only 1 study met all of the criteria for Item 4, which refers to the comprehensiveness of the search method.³⁹ Item 7, giving a list of excluded studies and the reason for excluding them, was also only seen in 1 study.^11,14

Table 2.

Results of the AMSTAR-2 Assessments

Abbreviations: Yes, Y; Partial Yes, PY; No, N.

Highlighted columns are the critical questions of AMSTAR-2 checklist.

Fig 2.

Radar chart of proportion of included studies meeting complete adherence to AMSTAR-2 items.

The radar chart showing the adherence rate of included studies to each item of AMSTAR-2 checklist.

Furthermore, although only 1 study showed full adherence to the criteria for using an optimal method for assessing the RoB of individual studies (item 9),²³ all other studies met the relevant criteria for a partial yes. Another item of the AMSTAR-2 tool that was observed in only 2 included reviews is reporting the sources of funding for the research included in the review (item 10).^21,23 Only items 11 and 13 were fully observed by more than half of the included research, which refer to applying proper statistical methods to integrate the results (item 11) and assessing the possibility of bias in the individual studies when interpreting. About 40% of included reviews and 30% of included reviews followed the statistical heterogeneity (item 14) and publication bias (small study effect) (item 15) guidelines, respectively. For further details, you can see Table 2 and Supplementary Figure 1.

Results of Reporting Quality Evaluation

The results of how well each of the included studies reported their results are shown in Supplementary Table 3, which is set up according to the PRISMA criteria. In addition, Figure 3 illustrates the percentage of total and partial compliance among the included studies with these criteria. According to the findings, only 3 of the criteria on this checklist—specifically, questions 4 (objectives), 17 (study characteristics), and 20b (results of syntheses)—were satisfied by all of the included reviews. Nonetheless, none of the included studies adhered to questions 21 (which asked about publication bias) and 24c (which asked about registering the protocol).

Fig 3.

PRISMA results of 18 included reviews. The PRISMA results, visually summarizes the percentage of total and partial compliance among the included studies

Concerning the method section, the sixth question, which is about the sources of information sought, was fully addressed by only 2 studies (11%). This was primarily due to the lack of a comprehensive search for other sources in order to find gray literature. Furthermore, just 3 of the reviews, which is only 17%, gave complete search strategies for the electronic databases that were used (Q7). Q13e and Q13f, which mention exploring potential causes of heterogeneity and sensitivity analysis, were considered in less than 33% of the included studies, despite being critical in SR studies. Conversely, more than 70% of the included studies met the requirements for certain aspects of the method section. These aspects include Q10a (listing and defining all outcomes), Q5 (specifying the inclusion and exclusion criteria), Q9 (specifying the techniques used to collect data), and Q8 (Specifying the methods used to decide whether a study met the inclusion criteria). For more details, you can see Figure 4, which demonstrates the proportions of review studies meeting total or partial compliance with PRISMA criteria with associated 95% CIs.

Fig 4.

Forest plot of the proportion of included studies meeting (A) complete adherence to PRISMA items and (B) at least partially adherence to PRISMA items. A forest plot is displayed showing the proportion of included studies that (A) fully adhered to PRISMA items and (B) those that only partially adhered to PRISMA items.

Results of Risk of Bias Evaluation

Figure 5 and Supplemental Table 4 show the results of the RoB assessment for each study using the ROBIS tool. Following this tool, in 11.1% of reviews (n = 2), no concern regarding RoB was observed in the domain of study eligibility criteria”, 27.8% of reviews (n = 5) in the domain of “identification and selection of studies”, 44.4% of reviews (n = 8) in the domain of “data collection and study appraisal”, and 27.8% of reviews (n = 5) in the domain of “synthesis and findings”. While 16.2% of the studies (n = 3) were rated as having a low RoB for domains 2 and 3,^13,23,25 22.2% of the included reviews (n = 4) received the most ratings that indicated they had a high RoB for these 2 areas.^14,16,17,28 Only in Domain 4, "synthesis and findings," did none of the included studies have a high RoB. According to the overall assessment carried out with the ROBIS tool, 16.2% (n = 3) of the studies were appraised as having a low RoB,^23,24,26 38.9% (n = 7) of the studies were appraised as having an unclear RoB,^{13, 14, 15,18,25,29,30} and 44.4% (n = 8) of the studies were appraised as having a high RoB.^{16,17,19, 20, 21, 22,27,28}

Fig 5.

ROBIS results of 18 included reviews. The ROBIS results showing the adherence rate of included studies to each item of ROBIS checklist.

Results of Effectiveness of ESWT

Table 3 summarizes the various outcomes investigated in studies assessing the effects of shock wave therapy.

Table 3.

Extracorporeal Shock Wave Therapy for Various Treatment Objectives

Outcomes	Percent of Studies	References
Pain	100%	13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30
Function/ disability	88.88%	13, 14, 15, 16, 17, 19, 20, 21, 22, 24, 25, 26, 27, 28, 29, 30
Disappearance of calcification	33.33%	14, 18, 20, 25, 29, 30
Satisfaction	5.55%	15
Quality of life	5.55%	15
Range of motion	5.55%	23
Treatment success rate	11.11%	23, 24

Upper Limb

Shoulder Tendinosis

Five studies^{14,20,22,30,40} with critically low quality assessed the efficacy of the ESWT for shoulder tendinosis treatment. Four out of 5 studies^14,20,30,40 concluded that high-energy ESWT was more effective than low-energy ESWT in reducing pain, improving function, and causing the disappearance of calcium deposits in patients with shoulder tendinosis. One study compared the effectiveness of ESWT treatment with a placebo or between different intensities of ESWT and reported consistent evidence for the midterm effectiveness of ESWT in reducing pain and improving shoulder function for patients with shoulder tendinosis.²²

Rotator Cuff Tendinitis

Six studies^{18,19,21,25,28,39} with critically low quality reported the pooled results of the efficacy of the ESWT for rotator cuff tendonitis treatment. The results of 2 studies^18,19 suggest that high-energy ESWT is more effective for calcific than non-calcific rotator cuff tendonitis. Also, they reported that there was no evidence for the effectiveness of low-energy ESWT compared to a placebo.

One study²¹ compared ESWT with ultrasound-guided lavage with a corticosteroid injection and showed that ultrasound-guided lavage with a corticosteroid injection is more effective than shockwave therapy in reducing pain and improving function. Another study³⁹ compared the efficacy of ESWT with traditional rehabilitation and found no consensus between the 2 interventions for reducing pain and improving function. The effectiveness of high-energy ESWT versus arthroscopic surgery in the management of calcific rotator cuff tendinopathy was assessed in another study.²⁵ Both interventions were effective in reducing pain, improving function, and causing the disappearance of calcium deposits.

One study compared the efficacy of the ESWT with a placebo and concluded that the ESWT was more effective than the placebo in the management of rotator cuff tendinopathies.²⁸

Lower Limb Tendinopathies

A total of 2 studies^7,26 with critically low quality compared to the effectiveness of ESWT with a placebo or any other treatment in patients with any type of lower limb tendinopathy. Compared with the placebo control, ESWT had significant effects on reducing pain and improving functional outcomes. ESWT and invasive control showed no difference in results.

Patellar Tendinopathy

Two studies^23,27 with critically low quality assessed the effectiveness of ESWT in the treatment of patients with patellar tendinopathy. One study²³ compared the efficacy of different ESWT types, energy levels, and intervention durations. The results revealed moderately significant evidence of the safety and efficacy of general ESWT in increasing the treatment success rate, pain reduction, and range of motion recovery. Another study²⁷ compared the effectiveness of ESWT with placebo or eccentric exercises in patients with patellar tendinopathy and reported no significant differences between the intervention and control groups in reducing pain and improving function.

Achilles Tendinopathy

A total of 3 studies,^13,16,27 all of the critically low quality, compared pain and functional outcomes between ESWT and placebo or other nonsurgical interventions. Two out of 3 studies^13,16 reported significantly better results for ESWT compared to placebo and other nonsurgical interventions, and the results of another study²⁷ indicated no significant differences between the intervention and control groups.

Discussion

The overall level of methodological quality of all of the SR studies that evaluated the effects of ESWT on tendinopathies was described as being "critically low." To be more specific, the included reviews commonly lacked a registered and published protocol, did not provide a list of excluded studies along with the reason for exclusion, did not describe the selection study designs for inclusion in the review, did not report the sources of funding for included studies in the review, did not assess the potential impact of RoB's included studies on the results, did not adequately explain the heterogeneity among the studies and potential reasons for such a condition, and did not properly investigate publication bias.

According to the included SRs/MAs, ESWT is potentially effective in terms of improvements in pain, function, disappearance of calcification, satisfaction, quality of life, range of motion, and treatment success rate of patients with upper and lower limb tendinopathy. The classification of ESWT varied between studies. Two studies^23,24 used a double distinction based on intensity levels (high and low) and wave types (radial and focused). There were superior pooled effects of high-dose-focused ESWT, high-dose radial ESWT, and low-dose radial ESWT for overall clinical outcomes. The findings of these studies regarding the higher impact of high-dose ESWT compared to low-dose ESWT are in line with the findings of SRs/MAs that compared the efficacy of ESWT based on intensity levels.^{14,18, 19, 20,29,30} According to the findings, it seems that high-dose ESWT is more effective than low-dose ESWT for the treatment of upper and lower limb tendinopathy.

A comparison of ESWT to placebo, no treatment, non-operative, and surgical interventions indicates that it may be more effective than placebo or no treatment and equal to nonoperative and surgical interventions for the treatment of tendinopathy. Only 1 study²¹ suggests that ultrasound-guided lavage with a corticosteroid injection is more effective than ESWT in the treatment of rotator cuff tendinopathy. As a result, ESWT can be considered an effective treatment for upper and lower limb tendinopathy.

The general confidence in the results of SRs is very important for giving unbiased information about how a treatment works, since clinicians and people who make clinical guidelines often use the results of this research. So, if clinicians want to use SR findings in clinical settings, they need to know a lot about how well the studies are planned and done.⁴¹ In this way, people have been talking about the methodological quality of SRs for the past 20 years, which led to the first version of the AMSTAR-2 tool being made in 2007.⁴² The term "methodological quality" refers to how well the planning and execution of a SR can lead to results that are accurate and fair.⁴³ Until now, various checklists have been developed to evaluate the methodological quality of SRs, including the AMSTAR-2 tool, in order to assist users in interpreting the findings of this type of study.

In our study, the methodological quality of all included reviews was evaluated as critically low. In accordance with this finding, Almeida et al.⁴⁴ evaluated the methodological quality of SRs assessing exercise therapy for chronic low back pain and showed that only 8% of included reviews were rated as high in accordance with the AMSTAR-2 tool. Furthermore, Oliveira et al. (2017) revealed that a major proportion of SRs in the field of physical therapy had low methodological quality based on the AMSTAR-2 items.⁴⁵ Such a situation can be observed in other fields of medicine as well. For instance, a recent review of research that evaluated the methodological quality of SRs in the field of bariatrics found that 99% of the included studies were rated as “critically low” using the AMSTAR-2 instrument.⁴⁶ The conducted studies in the fields of orthopedics, pediatric oncology, orthodontics, and diabetes mellitus also reported similar results.^{47, 48, 49, 50}

The methodological quality of the studies that were looked at is not good enough for a number of reasons. Based on our results, nearly all included reviews lacked any registered and published protocols, which is one of the critical items of the AMSTAR-2 tool. Albeit it should be noted that the PROSPERO system was launched in early 2011 in order to register the protocols of systematic review studies, and about 33% of our included studies were published before that. Two previous SRs in the field of physical therapy found that the proportion of included studies with a registered protocol was low (<21%), which is consistent with our findings.^44,45 In this regard, the relevant literature has shown that reviews with registered protocols commonly generate higher scores using the AMSTAR-2 tool.^45,51,52 Therefore, just as the protocol registration code is necessary for randomized controlled trials, journal editors should also request that review authors prospectively register their protocols in electronic databases such as PROSPERO. Pre-registration of the protocol can indeed reduce the probability of bias because the methods and statistical analyses are clearly stated, which will result in greater transparency of the study.⁵³

Even though almost all of the included reviews used an appropriate approach for figuring out how likely bias was in each study, only 10% of them thought about how bias might affect the results. Similar to what we found, a cross-sectional study showed that even though more than 80% of review studies used the right method to figure out the RoB, very few looked at how bias affected the results.⁴⁴ Individual studies with a high RoB may exaggerate the effects of interventions, so the conclusions of these SRs should be taken with care. In this regard, review study authors should know that assessing the RoB is not enough; they should also think about how the status of bias might affect the study results. It's important to keep in mind that the AMSTAR-2 tool evaluates the methodological quality of SRs, not primary research.³⁵ Because of this, a SR may contain primary studies that have a high RoB; yet, because the review was carried out effectively, it is considered to be of high quality.

Another important result that this study found was that only half of the review studies looked into what caused heterogeneity or how it affected the results of the study. Even though statistical heterogeneity may be found in meta-analyses with tools like the I² or Q-Cochrane test, the reasons for it and how it affects the results are rarely looked into.⁵⁴ This can lead to false-negative or false-positive results.⁵⁵ Because of this, the Cochrane Handbook suggests that researchers doing SR studies should take into account both statistical and methodological heterogeneity to get more reliable results.⁵³ Also, in our study, only 30% of the meta-analyses that were included looked at small study bias (also called publication bias) enough. Almeida et al. (2019) found that only 21% of systematic reviews on exercise therapy for chronic low back pain looked into this kind of bias, which is the same as what we found.⁴⁴ Because studies with positive results are published more often and faster than those with negative results,⁵⁶ evaluating this type of bias can make it less likely that the effects of the intervention were overestimated.

Based on the findings of the reporting quality assessment of included reviews using the PRISMA checklist version 2020, the highest adherence to items 4, 17, and 20b was observed, and the lowest adherence to items 21, 24c, 24a, and 24b. Given the overlap of the PRISMA checklist with the AMSTAR-2 tool, we will discuss only those that have not been addressed before. Findings from the reporting quality assessment showed that less than 30% of the reviews assessed the certainty (or confidence) in the body of evidence for an outcome. In recent years, a number of tools have been developed to assess the quality of the body of evidence, including GRADE (Grading of Recommendations Assessment, Development, and Evaluation).⁵⁷ This tool has some domains, including RoB assessment, inconsistency, indirectness, publication bias, and imprecision, which demonstrates the strength of the findings of the SR.⁵⁸ As a result, it is imperative that the GRADE tool be utilized in upcoming SRs in order to demonstrate the quality status of conclusive data in an effort to facilitate the process of guideline development.

ROBIS was used to figure out the RoB of each review, and only 16.16% of the reviews were found to have a low RoB. In this regard, Storman et al. (2019) revealed that just 6% of SRs published in the field of bariatrics were evaluated as low RoB.⁴⁶ Additionally, another study showed that only 3% of SRs addressing nutrition for cancer prevention had a low RoB based on the ROBIS.⁵⁹ Therefore, in order to conduct review studies with minimal bias, it is necessary for SRs to apply the domains presented in the ROBIS tool in their studies. Furthermore, it is highly recommended that users of review studies, editors, and guideline developers should use such tools to determine the RoB status of this type of study.

Strengths

This review study's strengths include a thorough search plan that is not limited to 1 language and independent screening and selection of the final studies. This study is the first in this field.

Limitations

Firstly, our findings are limited to SRs in the field of ESWT for subjects with tendinopathy. Subjectively involved in the quality evolution of included reviews was another limitation of the present study. The research team attempted to curb such subjectivity by conducting several calibration practices and providing clear instructions to authors regarding each of the items in our evaluation forms. Third, the data extraction and quality evaluation of the included reviews were not blinded, which could be a source of bias. In fact, we considered the officially full-text of articles through the study.

Conclusion

All included review studies had a critically low methodological quality according to the AMSTAR-2. The main reasons for this result were, among others, not following items 2 (pre-registered protocol) and 7 (reporting of excluded studies along with the reason they were left out). In addition, the reporting quality and RoB status of the included reviews were highly unsatisfactory. Therefore, SRs/MAs evaluating the effects of ESWT on tendinopathies should be interpreted with great caution.