ondasdechoque.org
PT-BR

Myofascial pain syndrome: multiple sites · Review · 2024

Does shockwave ease myofascial (trigger point) pain?

In this review of 27 trials, mostly of neck and shoulder pain, people treated with shockwave had about 1.7 points less pain on a 0–10 scale than with sham or control treatments, usually within 4 weeks. Against other treatments such as dry needling, injections or laser, the gap was about 0.4 points and uncertain.

Helped more than sham or control in the short term
trials
27
adults completed the trials
1,346
trials compared it with another treatment, such as dry needling or laser
18

Good to know: Most trials treated the neck and shoulders.

Avendaño-López C, Megía-García Á, Beltran-Alacreu H, et al. American Journal of Physical Medicine & Rehabilitation, 2024 · Read the paper ↗

How this study worked

10 studiesOne answer
A review gathers earlier studies to answer one question.

What happened

studies
10
Compared with
Sham or control treatments, and other treatments such as dry needling, ultrasound, injections, laser or exercise

Pain compared with other options

Difference on a 0–10 pain scale · below zero means less pain with shockwave

Pain was about 1.7 points lower with shockwave than with sham or control (10 trials). Against other treatments the average gap was about 0.4 points, with a range that includes no difference (18 trials).
Show the numbers
Difference (points on a 0–10 pain scale)likely range
Versus sham or control−1.66−2.18 to −1.14
Versus other treatments−0.36−0.73 to 0.01

What the trials used

Regimens in included trials
Radial ESWT in 18 trials (1.2–4 bar or 0.056–0.2 mJ/mm², 300–1,500 pulses on trigger points, 1–20 Hz) and focused ESWT in 9 (0.056–0.26 mJ/mm², 400–700 pulses, 1–10 Hz); one to three sessions per week, usually 3–4 in total (range 1–24). These are descriptions, not an established optimal regimen.
Comparators
Control or sham ESWT (12 trials) and other treatments (18 trials): dry needling, ultrasound, exercise, injections (including corticosteroid and lidocaine), electroacupuncture, laser and neuromuscular inhibition techniques; six trials had three arms.
Review methods
PRISMA, PROSPERO CRD42022344766; searches to 2 June 2022; English or Spanish RCTs; duplicate selection, extraction and RoB 2 assessment; inverse-variance random-effects pooling in RevMan 5.4.1 of the longest available assessment; GRADE.

Who took part

Adults aged 18–70 with trigger-point muscle pain, mostly in the neck and shoulders; a few trials treated the leg, foot or lower back.

Side effects

Most trials did not report side effects; those that did found none.

All the numbers and the full appraisal
FROM THE FULL PAPER

The results, in numbers.

27 randomized trials of adults with myofascial pain syndrome with trigger points (searched to 2 June 2022): 1,377 enrolled and 1,346 completed. The paper splits completers as 595 ESWT, 570 control/sham and 181 other treatment, but 595 and 570 equal the totals of the pain forest plot, where 570 includes 416 participants given other treatments; the split is therefore unreliable. 802 women where reported; ages 18–70. Twenty-two trials treated the neck and shoulder, three the leg or foot and two the lower back.

Pain intensity (VAS)2 reported outcomes
01
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Pain intensity (VAS)

VAS or numeric pain rating scale cm on a 10 cm VAS

MD-1.66CI 95% -2.18 to -1.14p <0.00001

ESWT versus control or sham ESWT

10 studies · I²=77%

Random-effects inverse-variance pooling

-2.4Equal results 00.3

Pain was lower with ESWT than with control by 1.66 cm on a 10 cm scale (95% CI −2.18 to −1.14), with high heterogeneity; the text rounds this to −1.7 (−2.2 to −1.1). The authors consider 1.5–3 cm clinically relevant. GRADE: low certainty (serious risk of bias and inconsistency).

Analysis details and source

10 comparisons; 162 ESWT and 154 control participants in the subtotal row. "Control" includes sham ESWT and other control conditions whose content is described only in the unread Supplement. Excluding trials without stated diagnostic criteria gave −1.7 (−2.3 to −1.0; I²=62%).

02
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Pain intensity (VAS)

VAS or numeric pain rating scale cm on a 10 cm VAS

MD-0.36CI 95% -0.73 to 0.01p = 0.06

ESWT versus other active treatments (pooled)

18 studies · I²=90%

Random-effects inverse-variance pooling

-0.8Equal results 00.1

Against other active treatments combined, pain differed by −0.36 cm (95% CI −0.73 to 0.01; P=0.06), with very high heterogeneity; individual comparisons ranged from clearly favouring ESWT to clearly favouring the comparator. GRADE: low certainty (serious risk of bias and inconsistency).

Analysis details and source

18 comparisons; 433 ESWT and 416 comparator participants in the subtotal row. Comparison rows are counted as printed. In three-armed trials the arm shared by two comparisons is split between rows, as the Methods state; in Cho 2012, Fang 2014 and Mohamed 2021 the ESWT column holds two different arms, and the same comparator arm is labelled "control" in one subgroup and "other" in the other, so the control/other grouping of these trials is unclear (Mohamed 2021 row 2 favours the comparator by 2.6 cm). Excluding trials without stated diagnostic criteria gave −0.2 (−0.6 to 0.2; I²=91%).

Pressure pain threshold2 reported outcomes
03
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Pressure pain threshold

Pressure algometry kg/cm²

MD1.07CI 95% 0.43 to 1.71p = 0.001

ESWT versus control or sham ESWT

10 studies · I²=97%

Random-effects inverse-variance pooling

-0.2Equal results 01.9

Pressure pain threshold was 1.07 kg/cm² higher with ESWT than control (95% CI 0.43 to 1.71), with very high heterogeneity. In subgroups, focused devices showed 1.8 (1.6 to 2.0) and radial 0.7 (0.04 to 1.4), subgroup difference P=0.003. GRADE: low certainty (serious risk of bias and inconsistency).

Analysis details and source

10 comparisons; 161 ESWT and 155 control participants. Excluding trials without stated diagnostic criteria gave 1.4 (1.0 to 1.9; I²=86%).

04
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Pressure pain threshold

Pressure algometry kg/cm²

MD0.12CI 95% -0.03 to 0.27p = 0.12

ESWT versus other active treatments (pooled)

14 studies · I²=86%

Random-effects inverse-variance pooling

-0.1Equal results 00.3

Against other active treatments, pressure pain threshold differed by 0.12 kg/cm² (95% CI −0.03 to 0.27). GRADE: low certainty (serious risk of bias and inconsistency).

Analysis details and source

14 comparisons; 289 ESWT and 287 comparator participants. Comparison rows are counted as printed. In three-armed trials the arm shared by two comparisons is split between rows, as the Methods state; in Cho 2012, Fang 2014 and Mohamed 2021 the ESWT column holds two different arms, and the same comparator arm is labelled "control" in one subgroup and "other" in the other, so the control/other grouping of these trials is unclear (Mohamed 2021 row 2 favours the comparator, MD −0.40 kg/cm²).

Region-specific function scales (e.g., NDI, ODI)2 reported outcomes
05
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Function

Region-specific function scales (e.g., NDI, ODI) Standard deviation units

SMD-0.83CI 95% -1.62 to -0.04p = 0.04

ESWT versus control or sham ESWT

6 studies · I²=83%

Random-effects inverse-variance pooling

-1.8Equal results 00.2

Function favoured ESWT over control by SMD −0.83 (95% CI −1.62 to −0.04), with high heterogeneity; one trial (Elhafez 2022, SMD −3.38) is far from the others. After excluding trials without stated diagnostic criteria the estimate was −0.9 (−2.0 to 0.2), an interval that includes no difference. GRADE: low certainty (serious risk of bias and inconsistency).

Analysis details and source

6 comparisons; 94 ESWT and 99 control participants. A standardized difference is not a point change on any function scale; negative favours ESWT.

06
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Function

Region-specific function scales (e.g., NDI, ODI) Standard deviation units

SMD-0.5CI 95% -1.06 to 0.05p = 0.08

ESWT versus other active treatments (pooled)

15 studies · I²=92% (Figure 4); the text reports 83%

Random-effects inverse-variance pooling

-1.2Equal results 00.2

Against other active treatments, function differed by SMD −0.50 (95% CI −1.06 to 0.05). GRADE: low certainty (serious risk of bias and inconsistency).

Analysis details and source

15 comparisons; 374 ESWT and 372 comparator participants. Heterogeneity is 92% in Figure 4 but 83% in the Results text. Comparison rows are counted as printed. In three-armed trials the arm shared by two comparisons is split between rows, as the Methods state; in Cho 2012 and Mohamed 2021 the ESWT column holds two different arms, and the same comparator arm is labelled "control" in one subgroup and "other" in the other, so the control/other grouping of these trials is unclear (Mohamed 2021 row 2 favours the comparator, SMD 2.77).

Pain intensity (VAS)7 reported outcomes
07
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Pain intensity (VAS)

cm on a 10 cm VAS

MD-0.1CI 95% -0.5 to 0.3

ESWT versus dry needling

4 studies · 190 participants in this estimate · I²=0%

Random-effects inverse-variance pooling

-0.6Equal results 00.4

ESWT versus dry needling: MD −0.1 (−0.5 to 0.3), with an interval that includes no difference; 4 studies, 190 participants, I² 0%.

Analysis details and source

From Table 2. Its footnote states that for pain and function negative values favour ESWT and for pressure pain threshold positive values favour ESWT. Sample size is the column printed in Table 2. Single-trial rows are one study, not a pooled estimate.

10
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Pain intensity (VAS)

cm on a 10 cm VAS

MD-1.4CI 95% -2.5 to -0.2

ESWT versus ultrasound

6 studies · 336 participants in this estimate · I²=91%

Random-effects inverse-variance pooling

-2.8Equal results 00.3

ESWT versus ultrasound: MD −1.4 (−2.5 to −0.2), favouring ESWT; 6 studies, 336 participants, I² 91%.

Analysis details and source

From Table 2. Its footnote states that for pain and function negative values favour ESWT and for pressure pain threshold positive values favour ESWT. Sample size is the column printed in Table 2. Single-trial rows are one study, not a pooled estimate.

13
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Pain intensity (VAS)

cm on a 10 cm VAS

MD-0.3CI 95% -0.7 to 0.1

ESWT versus injections (infiltration)

3 studies · 144 participants in this estimate · I²=89%

Random-effects inverse-variance pooling

-0.8Equal results 00.2

ESWT versus injections (infiltration): MD −0.3 (−0.7 to 0.1), with an interval that includes no difference; 3 studies, 144 participants, I² 89%.

Analysis details and source

From Table 2. Its footnote states that for pain and function negative values favour ESWT and for pressure pain threshold positive values favour ESWT. Sample size is the column printed in Table 2. Single-trial rows are one study, not a pooled estimate.

15
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Pain intensity (VAS)

cm on a 10 cm VAS

MD0.3CI 95% -0.6 to 1.3

ESWT versus laser

2 studies · 107 participants in this estimate · I²=0%

Random-effects inverse-variance pooling

-0.8Equal results 01.5

ESWT versus laser: MD 0.3 (−0.6 to 1.3), with an interval that includes no difference; 2 studies, 107 participants, I² 0%.

Analysis details and source

From Table 2. Its footnote states that for pain and function negative values favour ESWT and for pressure pain threshold positive values favour ESWT. Sample size is the column printed in Table 2. Single-trial rows are one study, not a pooled estimate.

17
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Pain intensity (VAS)

cm on a 10 cm VAS

MD-0.5CI 95% -1.4 to 0.4

ESWT versus exercise

1 study · 24 participants in this estimate · Not applicable (single study)

Random-effects inverse-variance pooling

-1.6Equal results 00.6

ESWT versus exercise: MD −0.5 (−1.4 to 0.4), with an interval that includes no difference; 1 study, 24 participants, I² not applicable (single study).

Analysis details and source

From Table 2. Its footnote states that for pain and function negative values favour ESWT and for pressure pain threshold positive values favour ESWT. Sample size is the column printed in Table 2. Single-trial rows are one study, not a pooled estimate.

18
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Pain intensity (VAS)

cm on a 10 cm VAS

MD-0.8CI 95% -1.5 to -0.1

ESWT versus electroacupuncture

1 study · 60 participants in this estimate · Not applicable (single study)

Random-effects inverse-variance pooling

-1.7Equal results 00.2

ESWT versus electroacupuncture: MD −0.8 (−1.5 to −0.1), favouring ESWT; 1 study, 60 participants, I² not applicable (single study).

Analysis details and source

From Table 2. Its footnote states that for pain and function negative values favour ESWT and for pressure pain threshold positive values favour ESWT. Sample size is the column printed in Table 2. Single-trial rows are one study, not a pooled estimate.

19
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Pain intensity (VAS)

cm on a 10 cm VAS

MD2.6CI 95% 2 to 3.2

ESWT versus neuromuscular inhibition techniques

1 study · 40 participants in this estimate · Not applicable (single study)

Random-effects inverse-variance pooling

-0.4Equal results 03.6

ESWT versus neuromuscular inhibition techniques: MD 2.6 (2.0 to 3.2), favouring neuromuscular inhibition techniques; 1 study, 40 participants, I² not applicable (single study).

Analysis details and source

From Table 2. Its footnote states that for pain and function negative values favour ESWT and for pressure pain threshold positive values favour ESWT. Sample size is the column printed in Table 2. Single-trial rows are one study, not a pooled estimate.

Pressure pain threshold2 reported outcomes
08
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Pressure pain threshold

kg/cm²

MD-0.1CI 95% -0.3 to 0.1

ESWT versus dry needling

4 studies · 188 participants in this estimate · I²=0%

Random-effects inverse-variance pooling

-0.3Equal results 00.1

ESWT versus dry needling: MD −0.1 (−0.3 to 0.1), with an interval that includes no difference; 4 studies, 188 participants, I² 0%.

Analysis details and source

From Table 2. Its footnote states that for pain and function negative values favour ESWT and for pressure pain threshold positive values favour ESWT. Sample size is the column printed in Table 2. Single-trial rows are one study, not a pooled estimate.

11
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Pressure pain threshold

kg/cm²

MD0.7CI 95% -0.4 to 1.7

ESWT versus ultrasound

3 studies · 146 participants in this estimate · I²=97%

Random-effects inverse-variance pooling

-0.7Equal results 02

ESWT versus ultrasound: MD 0.7 (−0.4 to 1.7), with an interval that includes no difference; 3 studies, 146 participants, I² 97%.

Analysis details and source

From Table 2. Its footnote states that for pain and function negative values favour ESWT and for pressure pain threshold positive values favour ESWT. Sample size is the column printed in Table 2. Single-trial rows are one study, not a pooled estimate.

Function4 reported outcomes
09
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Function

Standard deviation units

SMD-0.03CI 95% -0.4 to 0.3

ESWT versus dry needling

2 studies · 132 participants in this estimate · I²=6%

Random-effects inverse-variance pooling

-0.5Equal results 00.4

ESWT versus dry needling: SMD −0.03 (−0.4 to 0.3), with an interval that includes no difference; 2 studies, 132 participants, I² 6%.

Analysis details and source

From Table 2. Its footnote states that for pain and function negative values favour ESWT and for pressure pain threshold positive values favour ESWT. Sample size is the column printed in Table 2. Single-trial rows are one study, not a pooled estimate.

12
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Function

Standard deviation units

SMD-1CI 95% -1.6 to -0.4

ESWT versus ultrasound

6 studies · 336 participants in this estimate · I²=85%

Random-effects inverse-variance pooling

-1.8Equal results 00.2

ESWT versus ultrasound: SMD −1.0 (−1.6 to −0.4), favouring ESWT; 6 studies, 336 participants, I² 85%.

Analysis details and source

From Table 2. Its footnote states that for pain and function negative values favour ESWT and for pressure pain threshold positive values favour ESWT. Sample size is the column printed in Table 2. Single-trial rows are one study, not a pooled estimate.

14
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Function

Standard deviation units

SMD-2.2CI 95% -6.9 to 2.6

ESWT versus injections (infiltration)

2 studies · 114 participants in this estimate · I²=98%

Random-effects inverse-variance pooling

-8Equal results 03.7

ESWT versus injections (infiltration): SMD −2.2 (−6.9 to 2.6), with an interval that includes no difference; 2 studies, 114 participants, I² 98%.

Analysis details and source

From Table 2. Its footnote states that for pain and function negative values favour ESWT and for pressure pain threshold positive values favour ESWT. Sample size is the column printed in Table 2. Single-trial rows are one study, not a pooled estimate.

16
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Function

Standard deviation units

SMD0.1CI 95% -0.3 to 0.5

ESWT versus laser

2 studies · 107 participants in this estimate · I²=0%

Random-effects inverse-variance pooling

-0.4Equal results 00.6

ESWT versus laser: SMD 0.1 (−0.3 to 0.5), with an interval that includes no difference; 2 studies, 107 participants, I² 0%.

Analysis details and source

From Table 2. Its footnote states that for pain and function negative values favour ESWT and for pressure pain threshold positive values favour ESWT. Sample size is the column printed in Table 2. Single-trial rows are one study, not a pooled estimate.

Joint range of motion1 reported outcome
20
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Range of motion

Joint range of motion Standard deviation units

SMD9.15CI 95% 0.09 to 18.2p = 0.05

ESWT versus control or sham ESWT

2 studies · I²=96%

Random-effects inverse-variance pooling

-2.2Equal results 020.4

Two trials gave a pooled SMD of 9.15 (95% CI 0.09 to 18.20; text 9.2, 0.1 to 18.2) favouring ESWT over control. A standardized difference of this size is far outside usual ranges and reflects very small SDs in the two trials (e.g., Kamel 2020, SMD 13.88); it should not be read as a large functional gain. GRADE: low certainty (serious risk of bias and inconsistency).

Analysis details and source

2 comparisons; 37 ESWT and 36 control participants. Versus other treatments (3 comparisons) the SMD was −0.32 (−1.89 to 1.24).

Quality of life (SF-36)1 reported outcome
21
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Quality of life (SF-36)

Qualitative finding

Only versus other treatments (three trials: Aktürk 2018, Gezginaslan 2020, Király 2018; 99 ESWT and 96 comparator participants). Mean differences favoured ESWT for physical functioning 13.89 (9.93 to 17.86), role-physical 15.42 (8.96 to 21.88), role-emotional 21.67 (14.33 to 29.01, two estimable trials) and social functioning 10.68 (1.27 to 20.09); intervals for vitality, mental health, bodily pain and general health included no difference. Figure 5B also pools all eight domains into one total (11.56), although they come from the same people.

Analysis details and source

The pooled gains come mainly from Gezginaslan 2020; in Király 2018 the domain differences were smaller, with intervals that include no difference (e.g., physical functioning 8.10, −1.37 to 17.57). The Discussion attributes the QoL gain to comparison with ultrasound, but these three trials compared ESWT with laser, conventional physical therapy and an arm not described in the main text. GRADE: low certainty (serious risk of bias and inconsistency).

Subgroups versus control (dose, device, site)1 reported outcome
22
Longest assessment in each trial (4 weeks or less in most; the paper is inconsistent on whether any exceeded 12 weeks)

Subgroups versus control (dose, device, site)

Qualitative finding

Versus control, pain and function did not differ by dose, focused versus radial device or body site. Lower-dose trials (<1.5 bar or <0.1 mJ/mm²) showed VAS −2.3 (−3.0 to −1.7) versus −1.5 (−2.1 to −0.9) for higher doses, subgroup P=0.07. For pressure pain threshold, focused devices (1.8, 1.6 to 2.0) exceeded radial (0.7, 0.04 to 1.4), P=0.003, and cervical sites exceeded shoulder and leg, P=0.005.

Analysis details and source

Subgroups contain 1–7 trials each and are observational comparisons across trials, not randomized dose comparisons. The Discussion describes the low-dose cut-off as <0.08 mJ/mm², whereas Table 1 uses <0.1 mJ/mm² or <1.5 bar.

How treatment was delivered
Regimens in included trials

Radial ESWT in 18 trials (1.2–4 bar or 0.056–0.2 mJ/mm², 300–1,500 pulses on trigger points, 1–20 Hz) and focused ESWT in 9 (0.056–0.26 mJ/mm², 400–700 pulses, 1–10 Hz); one to three sessions per week, usually 3–4 in total (range 1–24). These are descriptions, not an established optimal regimen.

Comparators

Control or sham ESWT (12 trials) and other treatments (18 trials): dry needling, ultrasound, exercise, injections (including corticosteroid and lidocaine), electroacupuncture, laser and neuromuscular inhibition techniques; six trials had three arms.

Review methods

PRISMA, PROSPERO CRD42022344766; searches to 2 June 2022; English or Spanish RCTs; duplicate selection, extraction and RoB 2 assessment; inverse-variance random-effects pooling in RevMan 5.4.1 of the longest available assessment; GRADE.

HARMS & TOLERABILITY

Adverse effects were not analysed: 16 of 27 trials did not report them, and the trials that did (11 in Methods, 10 in Results) reported none. This is missing reporting, not evidence that ESWT causes no harm.

LIMITS OF THE ANALYSIS

Author-reported: heterogeneous doses and incomplete reporting of application parameters; different pain and function scales; short follow-up (under 12 weeks, 4 weeks or less in most trials). GRADE certainty was low for every outcome.

Reference and data issues: Seven trials named in the forest plots (for example Aktürk 2018, Ali 2016, Anwar 2022, Kamel 2020, Mohamed 2021) do not appear in the printed reference list, and reference numbers in the Results point to non-trial references; the full trial list is only in the unread Supplement. In Cho 2012, Fang 2014 and Mohamed 2021 the same comparator arm is labelled "control" in one subgroup and "other" in the other, against two different ESWT-containing arms, so the control/other grouping of these trials is unclear.

Inconsistent statements: Methods say 11 trials reported no adverse effects while Results say 10; Methods say follow-up subgroups were not possible because only one study had follow-up, while Results describe assessments at 10–24 weeks in 10 trials and the Discussion says none exceeded 12 weeks; the abstract lists five databases, Methods six (adding Scopus); Methods describe fixed-effect pooling when I² <50%, but all figures use random effects; function heterogeneity versus other treatments is 83% in text and 92% in Figure 4.

Extractor interpretation: the Conclusion that ESWT is effective "compared with control and ultrasound therapy" and the Discussion statement that it improves outcomes compared with electroacupuncture go beyond Table 2: versus ultrasound the pressure-threshold interval includes no difference, and electroacupuncture rests on one trial, favourable for pain but with a pressure-threshold interval that includes no difference (0.2; −0.3 to 0.7). The ROM SMD of 9.2 is implausibly large. "Control" groups are heterogeneous and only partly sham.

Disclosures: PROSPERO CRD42022344766; no funding source is stated; the authors report no conflicts of interest. Data available from the corresponding author on justified request.

Findings checked against the full source by a separate automated reviewer.

Limitations noted by the reviewers

  • Author-reported: heterogeneous doses and incomplete reporting of application parameters; different pain and function scales; short follow-up (under 12 weeks, 4 weeks or less in most trials). GRADE certainty was low for every outcome.
  • Reference and data issues: Seven trials named in the forest plots (for example Aktürk 2018, Ali 2016, Anwar 2022, Kamel 2020, Mohamed 2021) do not appear in the printed reference list, and reference numbers in the Results point to non-trial references; the full trial list is only in the unread Supplement. In Cho 2012, Fang 2014 and Mohamed 2021 the same comparator arm is labelled "control" in one subgroup and "other" in the other, against two different ESWT-containing arms, so the control/other grouping of these trials is unclear.
  • Inconsistent statements: Methods say 11 trials reported no adverse effects while Results say 10; Methods say follow-up subgroups were not possible because only one study had follow-up, while Results describe assessments at 10–24 weeks in 10 trials and the Discussion says none exceeded 12 weeks; the abstract lists five databases, Methods six (adding Scopus); Methods describe fixed-effect pooling when I² <50%, but all figures use random effects; function heterogeneity versus other treatments is 83% in text and 92% in Figure 4.
  • Extractor interpretation: the Conclusion that ESWT is effective "compared with control and ultrasound therapy" and the Discussion statement that it improves outcomes compared with electroacupuncture go beyond Table 2: versus ultrasound the pressure-threshold interval includes no difference, and electroacupuncture rests on one trial, favourable for pain but with a pressure-threshold interval that includes no difference (0.2; −0.3 to 0.7). The ROM SMD of 9.2 is implausibly large. "Control" groups are heterogeneous and only partly sham.
  • Disclosures: PROSPERO CRD42022344766; no funding source is stated; the authors report no conflicts of interest. Data available from the corresponding author on justified request.
Research review

Source-linked critical reading

Systematic review with meta-analysis

Review statusAI-assisted draft · independent automated checkThis automated check comes before the editor’s clinical review.
Design and evidence base

Systematic review and random-effects meta-analysis of 27 randomized trials (PRISMA, PROSPERO CRD42022344766), searched to 2 June 2022 in PubMed, Cochrane, CINAHL, PEDro, SPORTDiscus and Scopus; RoB 2 and GRADE were applied. Results were analysed separately against control or sham and against other treatments, using the longest assessment in each trial.

Population

27 randomized trials of adults with myofascial pain syndrome with trigger points (searched to 2 June 2022): 1,377 enrolled and 1,346 completed. The paper splits completers as 595 ESWT, 570 control/sham and 181 other treatment, but 595 and 570 equal the totals of the pain forest plot, where 570 includes 416 participants given other treatments; the split is therefore unreliable. 802 women where reported; ages 18–70. Twenty-two trials treated the neck and shoulder, three the leg or foot and two the lower back. Diagnostic criteria, symptom duration (14 trials over 3 months) and trial characteristics are detailed only in the unread Supplement.

Intervention or treatment network

Radial ESWT in 18 trials (1.2–4 bar or 0.056–0.2 mJ/mm², 300–1,500 pulses on trigger points, 1–20 Hz) and focused ESWT in 9 (0.056–0.26 mJ/mm², 400–700 pulses, 1–10 Hz); one to three sessions per week, usually 3–4 in total (range 1–24). These are descriptions, not an established optimal regimen.

Comparators

Control or sham ESWT (12 trials) and other treatments (18 trials): dry needling, ultrasound, exercise, injections (including corticosteroid and lidocaine), electroacupuncture, laser and neuromuscular inhibition techniques; six trials had three arms. The content of "control" arms is not described in the main text and is not uniformly sham.

Outcomes and timepoints

Primary: pain (VAS or NRS, cm), pressure pain threshold (kg/cm²) and region-specific function (SMD). Secondary: range of motion (SMD) and SF-36 quality of life; adverse effects were planned but not analysed. All but one trial assessed immediately after treatment; the longest assessment was 4 weeks or less in most trials, and the paper gives conflicting statements on follow-up beyond 12 weeks.

Comparative results

Versus control: pain MD −1.66 cm (95% CI −2.18 to −1.14; I² 77%), pressure threshold +1.07 kg/cm² (0.43 to 1.71; I² 97%), function SMD −0.83 (−1.62 to −0.04; I² 83%). Versus other treatments pooled: pain −0.36 (−0.73 to 0.01), threshold 0.12 (−0.03 to 0.27), function −0.50 (−1.06 to 0.05). By comparator, pain favoured ESWT over ultrasound (−1.4; −2.5 to −0.2) and in one electroacupuncture trial, showed no clear difference from dry needling (−0.1; −0.5 to 0.3), injections and laser, and favoured neuromuscular inhibition in one trial (2.6; 2.0 to 3.2). When trials without stated diagnostic criteria were excluded, function versus control became −0.9 (−2.0 to 0.2). All outcomes were rated low certainty.

Harms and tolerability

Adverse effects were not analysed: 16 of 27 trials did not report them, and the trials that did (11 in Methods, 10 in Results) reported none. This is missing reporting, not evidence that ESWT causes no harm.

Limitations and interpretation

Heterogeneous trials, doses and scales, high heterogeneity and short follow-up; low GRADE certainty throughout. Seven trials named in the forest plots (for example Aktürk 2018, Ali 2016, Anwar 2022, Kamel 2020, Mohamed 2021) do not appear in the printed reference list, and reference numbers in the Results point to non-trial references; the full trial list is only in the unread Supplement. In Cho 2012, Fang 2014 and Mohamed 2021 the same comparator arm is labelled "control" in one subgroup and "other" in the other, against two different ESWT-containing arms, so the control/other grouping of these trials is unclear. The Methods and Results also disagree on adverse-event reporting and follow-up, and the conclusion about ultrasound and electroacupuncture is broader than Table 2.

Applicability

Results apply mainly to neck and shoulder trigger-point pain treated for a few sessions and assessed soon after. They suggest a short-term pain benefit over control conditions and no clear average difference from dry needling, injections and laser; they do not identify a best device or dose, show lasting benefit, or establish safety.

What stronger research should do

Larger sham-controlled and head-to-head trials with stated diagnostic criteria, prespecified primary outcomes, follow-up beyond 3 months and structured harms reporting; a review that publishes its full trial list, labels comparators consistently and reports adverse events.

THE ORIGINAL SOURCE

Read the original article.

Efficacy of Extracorporeal Shockwave Therapy on Pain and Function in Myofascial Pain Syndrome: A Systematic Review and Meta-analysis of Randomized Clinical Trials.

Avendaño-López C, Megía-García Á, Beltran-Alacreu H, Serrano-Muñoz D, Arroyo-Fernández R, Comino-Suárez N, Avendaño-Coy J. · American journal of physical medicine & rehabilitation · 2024

Open the publication record or available full-text files.

Checked against the full paper · Clinical review: Dr. Marcus Yu Bin Pai, · How we review →