Greater trochanteric pain / gluteal tendinopathy · Review · 2023
How does shockwave rank among treatments for outer hip pain?
In this review of 8 studies (over 500 people), pain eased more with shockwave than with exercise alone, across 3 studies. Of six treatments, shockwave ranked second for pain relief, after platelet-rich plasma (PRP) injections. Against a steroid injection (one study), the gap was smaller and less certain.
Ranked second of six treatments for pain- studies in the review
- 8
- people with outer hip pain
- 500+
- treatments compared
- 6
Good to know: Rankings combine direct and indirect comparisons from 8 studies.
He Y, Lin Y, He X, et al. BMC Musculoskeletal Disorders, 2023 · Read the paper ↗
How this study worked
What happened
- studies
- 3
- Compared with
- Exercise alone, or an ultrasound-guided steroid injection
Pain: shockwave compared with exercise alone
Difference in pain score (VAS) · followed for up to 12 months
Show the numbers
| Difference (pain score points) | likely range | |
|---|---|---|
| Shockwave vs exercise alone | −3.16 | −3.64 to −2.68 |
What the trials used
- Shockwave intervention reporting
- Three studies compared ESWT with exercise and one with ultrasound-guided corticosteroid injection. The review does not provide ESWT device, focused/radial type, energy/pressure, pulses or session schedule in its study table.
Who took part
Over 500 people with outer hip pain in 8 studies of 24 to 120 people each, with average ages from about 49 to 64.
Side effects
The studies reported no complications other than pain where PRP was injected.
All the numbers and the full appraisal
The results, in numbers.
Eight controlled studies of greater trochanteric pain syndrome, 596 participants; sample sizes 24–120, mean ages 48.7–63.73 years, reported follow-up 24 weeks to 12 months. The total and the follow-up range are as printed in the text; the Table 1 sample sizes add up to 503, and Table 1 lists a 2-month follow-up for Ramon 2020.
Pain: PRP-U versus CSI-U7 reported outcomes
Pain: PRP-U versus CSI-U
Visual Analog Scale (VAS) VAS cm as reported in Table 1; meta-analysis score units
Ultrasound-guided platelet-rich plasma (PRP-U) versus ultrasound-guided corticosteroid injection (CSI-U)
2 studies · I²=58.3%
The estimate and interval are presented as published; scale direction depends on the stated comparison.
Analysis details and source
VAS range not explicitly specified as a native instrument bound; do not infer one. Follow-up varies across trials.
Pain: ESWT versus exercise
Visual Analog Scale (VAS) VAS cm as reported in Table 1; meta-analysis score units
Extracorporeal shock wave therapy (ESWT) versus exercise (EX)
3 studies · I²=27%
The estimate and interval are presented as published; scale direction depends on the stated comparison.
Analysis details and source
VAS range not explicitly specified as a native instrument bound; do not infer one. Follow-up varies across trials. Table 1 lists follow-up of 2 months (Ramon 2020), 12 months (Furia 2014) and 6 months (Shi 2021) for the three ESWT-versus-exercise trials, although the text gives a range of 24 weeks to 12 months. The Discussion reports this pairwise comparison as MD −3.15 (95% CI −4.10 to −2.17), unlike the Results and Table 4 (−3.16; −3.64 to −2.68).
Pain: ESWT versus CSI-U
Visual Analog Scale (VAS) VAS cm as reported in Table 1; meta-analysis score units
ESWT versus ultrasound-guided corticosteroid injection (CSI-U)
1 study
The estimate and interval are presented as published; scale direction depends on the stated comparison.
Analysis details and source
VAS range not explicitly specified as a native instrument bound; do not infer one. Follow-up varies across trials.
Pain: ultrasound-guided versus landmark CSI
Visual Analog Scale (VAS) VAS cm as reported in Table 1; meta-analysis score units
Ultrasound-guided CSI (CSI-U) versus landmark-guided CSI (CSI-B)
1 study
The estimate and interval are presented as published; scale direction depends on the stated comparison.
Analysis details and source
VAS range not explicitly specified as a native instrument bound; do not infer one. Follow-up varies across trials. The Results text prints MD 0.90 (95% CI −0.69 to 2.49) without saying which group was subtracted from which; Table 4 prints CSI-U vs CSI-B as −0.90 (−2.49, 0.69), and the Table 1 means for this trial (Mitchell 2018: CSI-U 1.3, CSI-B 2.2) point the same way as Table 4. The stored values follow the text.
Pain: CSI-U versus usual care
Visual Analog Scale (VAS) VAS cm as reported in Table 1; meta-analysis score units
Ultrasound-guided CSI (CSI-U) versus usual care (UC)
1 study
The estimate and interval are presented as published; scale direction depends on the stated comparison.
Analysis details and source
VAS range not explicitly specified as a native instrument bound; do not infer one. Follow-up varies across trials.
Network estimate: ESWT versus exercise
Visual Analog Scale (VAS) VAS cm as reported in Table 1; meta-analysis score units
Network estimate: ESWT versus exercise (EX)
3 studies
The estimate and interval are presented as published; scale direction depends on the stated comparison.
Analysis details and source
VAS range not explicitly specified as a native instrument bound; do not infer one. Follow-up varies across trials. Table 1 lists follow-up of 2 months (Ramon 2020), 12 months (Furia 2014) and 6 months (Shi 2021) for the three ESWT-versus-exercise trials, although the text gives a range of 24 weeks to 12 months. The paper labels the network-model intervals 95% CI (Abstract, Results, Table 4) and does not call them credible intervals.
Network estimate: PRP-U versus CSI-U
Visual Analog Scale (VAS) VAS cm as reported in Table 1; meta-analysis score units
Network estimate: PRP-U versus CSI-U
2 studies
The estimate and interval are presented as published; scale direction depends on the stated comparison.
Analysis details and source
VAS range not explicitly specified as a native instrument bound; do not infer one. Follow-up varies across trials. The paper labels the network-model intervals 95% CI (Abstract, Table 4) and does not call them credible intervals.
Treatment ranking probabilities (rank 1 worst, rank 6 best)1 reported outcome
Treatment ranking probabilities (rank 1 worst, rank 6 best)
VAS rank probability Probability
Table 3 assigns rank-6 probability 0.99 to PRP-U and rank-5 probability 0.81 to ESWT; exercise has rank-1 probability 0.84. These are ranking probabilities, not treatment effect sizes or certainty.
Analysis details and source
Rank 1 is worst and rank 6 best in the table note. Different treatment network and indirect comparisons limit interpretation.
How treatment was delivered
Three studies compared ESWT with exercise and one with ultrasound-guided corticosteroid injection. The review does not provide ESWT device, focused/radial type, energy/pressure, pulses or session schedule in its study table.
The discussion states no included studies reported complications other than pain at the injection site; systematic ascertainment and event counts are not supplied.
The authors report few studies, small samples, limited power and only VAS-including studies; evidence certainty ranged very low to moderate.
The article contains direction/number inconsistencies: prose gives CSI-U versus CSI-B MD 0.90 (−0.69 to 2.49), while Table 4 gives −0.90 (−2.49 to 0.69); the discussion reports PRP versus ESWT MD 4.37 with a wholly negative CI (−6.31 to −1.89), inconsistent with direction and network Table 2. These were not silently reconciled.
Supplementary files with pairwise meta-analysis detail were not acquired; estimates could not be checked against primary study data.
Findings checked against the full source by a separate automated reviewer.
Limitations noted by the reviewers
- The authors report few studies, small samples, limited power and only VAS-including studies; evidence certainty ranged very low to moderate.
- The article contains direction/number inconsistencies: prose gives CSI-U versus CSI-B MD 0.90 (−0.69 to 2.49), while Table 4 gives −0.90 (−2.49 to 0.69); the discussion reports PRP versus ESWT MD 4.37 with a wholly negative CI (−6.31 to −1.89), inconsistent with direction and network Table 2. These were not silently reconciled.
- Supplementary files with pairwise meta-analysis detail were not acquired; estimates could not be checked against primary study data.
Source-linked critical reading
Comparative evidence synthesis with meta-analysis
Technical source record
Complete XML including methods, all four raw tables, all three figure captions, results/discussion, limitations and declarations; primary trials and supplemental files unavailable.
The review found a possible pain-score advantage for ESWT over exercise alone, while direct evidence did not show a clear difference between ESWT and ultrasound-guided corticosteroid injection. The network favored PRP-U and ESWT rankings, but the evidence is limited and does not identify a proven best protocol.
Which conservative treatments reduce pain measured by the visual analog scale (VAS) in greater trochanteric pain syndrome (GTPS), and how do direct and network comparisons rank these options?
Bayesian network meta-analysis with consistency modeling, alongside DerSimonian–Laird random-effects pairwise meta-analyses. The authors searched PubMed, Cochrane Library and Web of Science through July 18, 2022; the protocol was registered as INPLASY 202280068. Eight controlled studies were included.
The synthesis included 596 participants across eight studies (as stated in the text; the Table 1 sample sizes add up to 503); individual samples ranged from 24 to 120, mean ages from 48.7 to 63.73 years, and follow-up from 24 weeks to 12 months (Table 1 lists 2 months for one ESWT-versus-exercise trial). All included studies measured VAS pain.
The review includes ESWT in three studies versus exercise and one versus ultrasound-guided corticosteroid injection (CSI-U). It does not report the ESWT device, focused/radial type, energy, pulse count or schedule in its study table, so a treatment protocol cannot be reconstructed from this review alone. Other network interventions were ultrasound-guided platelet-rich plasma (PRP-U), CSI-U, landmark-guided CSI (CSI-B), exercise and usual care.
The six-node network combines direct and indirect evidence across PRP-U, ESWT, CSI-U, CSI-B, exercise and usual care. Study counts belong to particular comparisons; they should not be summed as unique participants or treated as direct head-to-head evidence where no such trial existed.
All effects concern VAS pain. Pairwise estimates favored PRP-U over CSI-U (MD −5.02, 95% CI −6.25 to −3.79; 2 studies; I²=58.3%) and ESWT over exercise (MD −3.16, 95% CI −3.64 to −2.68; 3 studies; I²=27%). Pairwise comparisons were inconclusive for ESWT versus CSI-U (−0.82, −1.81 to 0.17), CSI-U versus CSI-B (prose 0.90, −0.69 to 2.49; Table 4 reverses the sign), and CSI-U versus usual care (−0.28, −0.94 to 0.37). The NMA estimated ESWT versus exercise at −3.17 (95% CI −4.13 to −2.15) and PRP-U versus CSI-U at −5.21 (−6.24 to −3.64). Table 3 assigns rank-6 probability 0.99 to PRP-U and rank-5 probability 0.81 to ESWT; exercise has rank-1 probability 0.84. These are ranking probabilities, not effect magnitudes.
The authors report an NMA advantage for ESWT versus exercise and PRP-U versus CSI-U, while direct ESWT versus CSI-U did not exclude no difference. The ranking analysis places PRP-U highest and ESWT next, but certainty is only very low to moderate and comparisons include indirect evidence. Source discrepancies remain: the discussion gives ESWT-versus-exercise MD −3.15 (95% CI −4.10 to −2.17), differing from Results/Table 4 (−3.16, −3.64 to −2.68), and gives PRP-versus-ESWT MD 4.37 with an entirely negative interval (−6.31 to −1.89). Do not silently reconcile these values.
The discussion states that no included study reported complications other than injection-site pain. It provides no event counts or systematic comparative ESWT-harm assessment; this limited reporting cannot establish that ESWT or injections are harmless.
Only eight small studies informed the network, fewer than ten studies per comparison precluded funnel-plot assessment, and risk-of-bias concerns included incomplete reporting of allocation concealment and participant/personnel blinding; one study had >10% loss to follow-up and four had samples below 50. Evidence certainty ranged from very low to moderate. Supplemental files containing detailed pairwise analyses were unavailable, and the paper has unresolved direction/value inconsistencies for CSI-U versus CSI-B, ESWT versus exercise and PRP versus ESWT.
This review supports possible VAS pain reduction with ESWT versus exercise, but it does not establish an optimal ESWT dose, a durable functional benefit, or superiority over CSI-U in direct evidence. The ranking is sensitive to a small, heterogeneous network and should not be presented as a definitive treatment hierarchy.
A sufficiently powered, preregistered multicenter trial should compare a fully specified ESWT protocol with credible active care and/or sham, use consistent VAS direction and clinically interpretable follow-up, report function and harms, and publish complete data. Future syntheses should reconcile comparison orientation and provide network assumptions, direct/indirect contribution and uncertainty transparently.
Read the original article.
The conservative management for improving Visual Analog Scale (VAS) pain scoring in greater trochanteric pain syndrome: a Bayesian analysis.
He Y, Lin Y, He X, Li C, Lu Q, He J. · BMC musculoskeletal disorders · 2023
Open the publication record or available full-text files.
Checked against the full paper · Clinical review: Dr. Marcus Yu Bin Pai, · How we review →