Trapezius pain: shockwaves versus injection and TENS · Randomized trial · 2012
Does shockwave help neck and shoulder muscle pain compared with injections?
In this trial, pain in the muscle across the top of the shoulder fell from about 7 to about 2 out of 10 after three shockwave sessions, one a week. With trigger-point injections plus TENS (mild electrical stimulation), it fell to about 3. Sore spots became less tender to pressure with shockwave.
Both eased pain- people with neck and shoulder muscle pain
- 30
- sessions, one a week
- 3
- week after the last session: final check
- 1
Good to know: All were hospital inpatients.
Jeon JH, Jung YJ, Lee JY, et al. Ann Rehabil Med, 2012 · Read the paper ↗
How this study worked
What happened
Neck and shoulder muscle pain
0–10 pain scale · start and 1 week after the third session
Start1 week after the last session
Show the numbers
| Group | People | Start | 1 week after the last session |
|---|---|---|---|
| Shockwave | 15 | 6.86 | 1.86 |
| Injections + TENS | 15 | 7.2 | 2.8 |
The treatment
Shockwave
Shockwave, once a week for 3 weeks
- Sessions
- 3, one a week
- Pulses
- 1,500 per session · 240 shocks a minute, aimed at the sore spots
- Strength
- 0.10mJ/mm²
- Numbing
- Not mentioned
- Device
- EvoTron RFL0300
Compared with: Trigger-point injections once a week plus TENS (mild electrical stimulation) 20 minutes a day, five days a week, for three weeks
Settings are what each study used, not a prescription.
Who took part
30 hospital patients (22 men, 8 women; average age 41–45) with pain in the muscle across the top of the shoulder and back of the neck.
Side effects
The study does not report side effects.
All the numbers and the full appraisal
The results, in numbers.
Hospital patients with posterior neck/shoulder pain and trapezius myofascial pain diagnosed on examination in 2011; 22 men and 8 women. Mean age was 40.86 in ESWT and 45.00 in control.
Source reports this participant group.
Source reports this participant group.
Pain intensity2 reported outcomes
Pain intensity
VAS points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| ESWT | 15 | 6.86 (SD 0.9) | 2.86 (SD 0.9) |
| Injection plus TENS | 15 | 7.2 (SD 1.3) | 4.6 (SD 0.55) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Reported group means; within-group change is not a between-group treatment effect.
Analysis details and source
Mean ± SD as printed. No effect or confidence interval reconstructed. Extractor's arithmetic check (not reported by the authors): every ESWT mean in this table is an exact multiple of 1/7 (e.g. 6.86 = 48/7, 1.86 = 13/7) and every control mean a multiple of 1/5, and the printed F and p values for this outcome fit 12 patients in total (group F = 9.940 with p = 0.010). These results are therefore consistent with about 7 ESWT and 5 control patients, not the 15 per group the paper reports; the paper does not explain this. Values and n are kept as printed.
Pain intensity
VAS points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| ESWT | 15 | 6.86 (SD 0.9) | 1.86 (SD 0.69) |
| Injection plus TENS | 15 | 7.2 (SD 1.3) | 2.8 (SD 0.84) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Reported group means; within-group change is not a between-group treatment effect.
Analysis details and source
Mean ± SD as printed. No effect or confidence interval reconstructed. Extractor's arithmetic check (not reported by the authors): every ESWT mean in this table is an exact multiple of 1/7 (e.g. 6.86 = 48/7, 1.86 = 13/7) and every control mean a multiple of 1/5, and the printed F and p values for this outcome fit 12 patients in total (group F = 9.940 with p = 0.010). These results are therefore consistent with about 7 ESWT and 5 control patients, not the 15 per group the paper reports; the paper does not explain this. Values and n are kept as printed.
Algometry2 reported outcomes
Pain threshold
Algometry lb/cm²
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| ESWT | 15 | 6.86 (SD 1.35) | 11.43 (SD 0.27) |
| Injection plus TENS | 15 | 6.2 (SD 1.92) | 8.8 (SD 0.48) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Reported group means; within-group change is not a between-group treatment effect.
Analysis details and source
Mean ± SD as printed. No effect or confidence interval reconstructed. Extractor's arithmetic check (not reported by the authors): every ESWT mean in this table is an exact multiple of 1/7 (e.g. 11.43 = 80/7, 12.57 = 88/7) and every control mean a multiple of 1/5, and its printed group and time F and p values fit 12 patients in total (group F = 5.237 with p = 0.045). These results are therefore consistent with about 7 ESWT and 5 control patients, not the 15 per group the paper reports; the paper does not explain this. Values and n are kept as printed. Table 3 prints the group×time test as F = 4.550 with p = 0.001 (the text repeats p = 0.001); an F of 4.550 on 2 numerator degrees of freedom corresponds to p of roughly 0.01–0.04, so the F or the p value is misprinted.
Pain threshold
Algometry lb/cm²
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| ESWT | 15 | 6.86 (SD 1.35) | 12.57 (SD 0.72) |
| Injection plus TENS | 15 | 6.2 (SD 1.92) | 9.6 (SD 2.19) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Reported group means; within-group change is not a between-group treatment effect.
Analysis details and source
Mean ± SD as printed. No effect or confidence interval reconstructed. Extractor's arithmetic check (not reported by the authors): every ESWT mean in this table is an exact multiple of 1/7 (e.g. 11.43 = 80/7, 12.57 = 88/7) and every control mean a multiple of 1/5, and its printed group and time F and p values fit 12 patients in total (group F = 5.237 with p = 0.045). These results are therefore consistent with about 7 ESWT and 5 control patients, not the 15 per group the paper reports; the paper does not explain this. Values and n are kept as printed. Table 3 prints the group×time test as F = 4.550 with p = 0.001 (the text repeats p = 0.001); an F of 4.550 on 2 numerator degrees of freedom corresponds to p of roughly 0.01–0.04, so the F or the p value is misprinted.
McGill pain2 reported outcomes
McGill pain
McGill points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| ESWT | 15 | 31 (SD 8.41) | 15.71 (SD 8.67) |
| Injection plus TENS | 15 | 24.4 (SD 6.02) | 16.8 (SD 5.17) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Reported group means; within-group change is not a between-group treatment effect.
Analysis details and source
Mean ± SD as printed. No effect or confidence interval reconstructed. Extractor's arithmetic check (not reported by the authors): every ESWT mean in this table is an exact multiple of 1/7 (e.g. 15.71 = 110/7, 8.71 = 61/7) and every control mean a multiple of 1/5, and the printed F and p values for this outcome fit 12 patients in total (group×time F = 4.150 with p = 0.031). These results are therefore consistent with about 7 ESWT and 5 control patients, not the 15 per group the paper reports; the paper does not explain this. Values and n are kept as printed.
McGill pain
McGill points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| ESWT | 15 | 31 (SD 8.41) | 8.71 (SD 2.81) |
| Injection plus TENS | 15 | 24.4 (SD 6.02) | 12 (SD 3.39) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Reported group means; within-group change is not a between-group treatment effect.
Analysis details and source
Mean ± SD as printed. No effect or confidence interval reconstructed. Extractor's arithmetic check (not reported by the authors): every ESWT mean in this table is an exact multiple of 1/7 (e.g. 15.71 = 110/7, 8.71 = 61/7) and every control mean a multiple of 1/5, and the printed F and p values for this outcome fit 12 patients in total (group×time F = 4.150 with p = 0.031). These results are therefore consistent with about 7 ESWT and 5 control patients, not the 15 per group the paper reports; the paper does not explain this. Values and n are kept as printed.
Pain rating2 reported outcomes
Pain rating
PRS points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| ESWT | 15 | 26.57 (SD 18.54) | 7.86 (SD 5.52) |
| Injection plus TENS | 15 | 25.2 (SD 7.95) | 11.2 (SD 2.59) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Reported group means; within-group change is not a between-group treatment effect.
Analysis details and source
Mean ± SD as printed. No effect or confidence interval reconstructed. Extractor's arithmetic check (not reported by the authors): every ESWT mean in this table is an exact multiple of 1/7 (e.g. 7.86 = 55/7, 4.57 = 32/7) and every control mean a multiple of 1/5, and the printed F and p values for this outcome fit 12 patients in total (time F = 10.864 with p = 0.004, Wilks' lambda). These results are therefore consistent with about 7 ESWT and 5 control patients, not the 15 per group the paper reports; the paper does not explain this. Values and n are kept as printed.
Pain rating
PRS points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| ESWT | 15 | 26.57 (SD 18.54) | 4.57 (SD 3.6) |
| Injection plus TENS | 15 | 25.2 (SD 7.95) | 8.6 (SD 0.89) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Reported group means; within-group change is not a between-group treatment effect.
Analysis details and source
Mean ± SD as printed. No effect or confidence interval reconstructed. Extractor's arithmetic check (not reported by the authors): every ESWT mean in this table is an exact multiple of 1/7 (e.g. 7.86 = 55/7, 4.57 = 32/7) and every control mean a multiple of 1/5, and the printed F and p values for this outcome fit 12 patients in total (time F = 10.864 with p = 0.004, Wilks' lambda). These results are therefore consistent with about 7 ESWT and 5 control patients, not the 15 per group the paper reports; the paper does not explain this. Values and n are kept as printed.
Total ROM2 reported outcomes
Total neck motion
Total ROM degrees
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| ESWT | 15 | 267.86 (SD 17.29) | 297.86 (SD 12.86) |
| Injection plus TENS | 15 | 263 (SD 31.35) | 290 (SD 20.62) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Reported group means; within-group change is not a between-group treatment effect.
Analysis details and source
Mean ± SD as printed. No effect or confidence interval reconstructed. Extractor's arithmetic check (not reported by the authors): every ESWT mean in this table is an exact multiple of 1/7 (e.g. 267.86 = 1,875/7, 302.14 = 2,115/7) and every control mean a multiple of 1/5, and the printed F and p values for this outcome fit 12 patients in total (group F = 0.379 with p = 0.552). These results are therefore consistent with about 7 ESWT and 5 control patients, not the 15 per group the paper reports; the paper does not explain this. Values and n are kept as printed. One other Table 5 pair (left-rotation group, F = 0.496 with p = 0.186) fits no sample size. Table 2 prints the control baseline as 263.00 ± 31.34; Table 5 prints ± 31.35 (stored).
Total neck motion
Total ROM degrees
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| ESWT | 15 | 267.86 (SD 17.29) | 302.14 (SD 9.94) |
| Injection plus TENS | 15 | 263 (SD 31.35) | 298 (SD 7.58) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Reported group means; within-group change is not a between-group treatment effect.
Analysis details and source
Mean ± SD as printed. No effect or confidence interval reconstructed. Extractor's arithmetic check (not reported by the authors): every ESWT mean in this table is an exact multiple of 1/7 (e.g. 267.86 = 1,875/7, 302.14 = 2,115/7) and every control mean a multiple of 1/5, and the printed F and p values for this outcome fit 12 patients in total (group F = 0.379 with p = 0.552). These results are therefore consistent with about 7 ESWT and 5 control patients, not the 15 per group the paper reports; the paper does not explain this. Values and n are kept as printed. One other Table 5 pair (left-rotation group, F = 0.496 with p = 0.186) fits no sample size. Table 2 prints the control baseline as 263.00 ± 31.34; Table 5 prints ± 31.35 (stored).
Global response and motion1 reported outcome
Global response and motion
Qualitative finding
Roles/Maudsley: excellent/good 6/9 with ESWT; excellent/good/acceptable 3/9/3 with control (p=.233). Trigger-point response excellent/good 10/5 versus 6/9 (p=.272). All neck-motion group and interaction tests were nonsignificant.
Analysis details and source
Qualitative extraction preserves the source scope.
How treatment was delivered
EvoTron RFL0300 electrohydraulic low-energy ESWT, 1,500 shocks per session, 0.10 mJ/mm² and 240 shocks/minute; three weekly sessions (4,500 total shocks), targeting twitch/referred-pain sites.
Trigger-point injections once weekly for three weeks plus TENS 20 minutes/day, five times weekly for three weeks. Injection drug and dose are not reported sufficiently for replication.
The paper discusses expected safety of low-energy ESWT but does not provide systematic event counts for this trial. Absence of a harm table is not evidence of zero adverse events.
Hospital sample and brief observation, unclear allocation safeguards, multiple outcomes and incomplete comparator details. The distinction between group main effect and interaction is essential: authors overinterpret VAS superiority despite a null interaction. No dedicated funding/conflict statement was found in the supplied article.
Extractor's arithmetic check (not reported by the authors): the continuous results in Tables 2–6 fit about 12 analysed patients rather than 30. Every ESWT mean is an exact multiple of 1/7 and every control mean a multiple of 1/5; Table 6's means for 'n=22' men and 'n=8' women are multiples of 1/9 and 1/3; and the printed F and p values fit 12 patients in total. The categorical counts (Table 1, Roles–Maudsley, final response) do cover 15 per group. Two printed F and p pairs do not match each other at any sample size: pain-threshold group×time (F = 4.550, p = 0.001; that F gives p of roughly 0.01–0.04) and left-rotation group (F = 0.496, p = 0.186). The paper explains none of this.
Findings checked against the full source by a separate automated reviewer.
Limitations noted by the reviewers
- Hospital sample and brief observation, unclear allocation safeguards, multiple outcomes and incomplete comparator details. The distinction between group main effect and interaction is essential: authors overinterpret VAS superiority despite a null interaction. No dedicated funding/conflict statement was found in the supplied article.
- Extractor's arithmetic check (not reported by the authors): the continuous results in Tables 2–6 fit about 12 analysed patients rather than 30. Every ESWT mean is an exact multiple of 1/7 and every control mean a multiple of 1/5; Table 6's means for 'n=22' men and 'n=8' women are multiples of 1/9 and 1/3; and the printed F and p values fit 12 patients in total. The categorical counts (Table 1, Roles–Maudsley, final response) do cover 15 per group. Two printed F and p pairs do not match each other at any sample size: pain-threshold group×time (F = 4.550, p = 0.001; that F gives p of roughly 0.01–0.04) and left-rotation group (F = 0.496, p = 0.186). The paper explains none of this.
Source-linked critical reading
randomized trial
Technical source record
Complete local PDF text read, including all article body sections, tables, discussion and references. The article has no figures, lists no supplementary material and has no funding or conflict-of-interest statement.
Pain scores improved in both groups. Some pressure-pain findings favored shockwaves, but this short-term trial does not establish a lasting advantage.
Can ESWT improve trapezius myofascial pain compared with trigger-point injection plus TENS?
Prospective randomized experimental/control study with 30 inpatients, 15 per group. No sequence-generation, concealment or masking procedure is described.
Hospital patients with posterior neck/shoulder pain and trapezius myofascial pain diagnosed on examination in 2011; 22 men and 8 women. Mean age was 40.86 in ESWT and 45.00 in control.
EvoTron RFL0300 electrohydraulic low-energy ESWT, 1,500 shocks per session, 0.10 mJ/mm² and 240 shocks/minute; three weekly sessions (4,500 total shocks), targeting twitch/referred-pain sites.
Trigger-point injections once weekly for three weeks plus TENS 20 minutes/day, five times weekly for three weeks. Injection drug and dose are not reported sufficiently for replication.
Algometric pain threshold, VAS, McGill pain questionnaire, pain rating scale, neck motion, Roles/Maudsley and trigger-point response. Assessments were before treatment, one week after first session and one week after third session.
Final VAS was 1.86±0.69 vs 2.80±0.84; threshold 12.57±0.72 vs 9.60±2.19 lb/cm² (ESWT/control). Group effects were p=.010 and .045, but VAS group×time was not significant (.162), whereas threshold interaction was printed as .001 (its F of 4.550 implies roughly .01–.04). McGill and pain-rating group effects were null (.816/.644); McGill interaction was .031. Total neck motion and global-response comparisons were not significant.
The paper discusses expected safety of low-energy ESWT but does not provide systematic event counts for this trial. Absence of a harm table is not evidence of zero adverse events.
Hospital sample and brief observation, unclear allocation safeguards, multiple outcomes and incomplete comparator details. The distinction between group main effect and interaction is essential: authors overinterpret VAS superiority despite a null interaction. No dedicated funding/conflict statement was found in the supplied article. The printed continuous results are arithmetically consistent with about 12 analysed patients (7 ESWT, 5 control) rather than 30, which the paper does not explain.
Findings concern hospitalized trapezius-pain patients and this low-energy protocol; they do not establish durable recovery, an optimal dose, or equivalence to injections/TENS.
Concealed randomized trials should prespecify a primary contrast, quantify harms, standardize injection/TENS and follow patients beyond the immediate course.
Read the original article.
The Effect of Extracorporeal Shock Wave Therapy on Myofascial Pain Syndrome
Jeon JH, Jung YJ, Lee JY, Choi JS, Mun JH, Park WY, Seo CH, Jang KU. · Ann Rehabil Med · 2012
Open the publication record or available full-text files.
Checked against the full paper · Clinical review: Dr. Marcus Yu Bin Pai, · How we review →