Neck and trapezius myofascial pain · Randomized trial · 2025
Does adding radial shockwave to exercises help neck and upper-back muscle pain?
With radial shockwave plus exercises and advice, pain fell from about 7 to just under 5 out of 10 by week 12. People who had a sham with the same exercises improved as much or more, to about 3.5.
- adults with neck or upper-back muscle pain
- 70
- shockwave sessions, one a week
- 6
- weeks of follow-up
- 12
Good to know: Both groups did the same exercise programme and got the same advice. About half were seen at week 12; for the others, their last score was used.
Ogbeivor C, AlMubarak H, Akomolafe T, et al. BMC Musculoskeletal Disorders, 2025 · Read the paper ↗
How this study worked
What happened
Neck and upper-back pain
0–10 pain scale · start to week 12
Shockwave + exerciseSham + exercise
Show the numbers
| Group | Start | Week 4 | Week 8 | Week 12 |
|---|---|---|---|---|
| Shockwave + exercise | 6.9 | 4.9 | 4.7 | 4.6 |
| Sham + exercise | 7.1 | 4.7 | 4.5 | 3.5 |
The treatment
Shockwave + exercise
Radial, once a week for 6 weeks
- Sessions
- 6, one a week
- Pulses
- 2,000 per session
- Strength
- 1.5bar (the paper also gives two different energy values: 0.57 and 0.068 mJ/mm²) · 15 per second
- Numbing
- Not mentioned
- Device
- Storz Medical
- Everyone also got
- Education, stretches and exercises, plus a home exercise plan
Compared with: A low-energy sham (0.3 bar) with the same exercise care
Settings are what each study used, not a prescription.
Who took part
70 adults with neck or upper-back muscle pain and tender trigger points, treated at one outpatient rehabilitation department in Riyadh.
Side effects
Two people in the shockwave group had brief extra tenderness and kept going. The paper also says one person stopped treatment at week 4 because pain increased. No serious side effects.
All the numbers and the full appraisal
The results, in numbers.
Adults aged ≥19 with neck and/or upper-back myofascial pain, palpable tenderness and active trigger point(s), recruited in one Riyadh outpatient rehabilitation department.
Authors state ITT with last observation carried forward; 20/70 were lost by week 4. Fig. 1 gives the numbers seen at each visit (shockwave 26, 18 and 15; placebo 23, 12 and 18 at weeks 4, 8 and 12); all 34 and 36 were analysed with carried-forward values.
Authors state ITT with last observation carried forward; 20/70 were lost by week 4. Fig. 1 gives the numbers seen at each visit (shockwave 26, 18 and 15; placebo 23, 12 and 18 at weeks 4, 8 and 12); all 34 and 36 were analysed with carried-forward values.
Numeric Pain Score (NPS)3 reported outcomes
Numeric Pain Score (NPS)
Numeric Pain Score (NPS) points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Low-energy sham plus shared care | 36 | 7.1 (SD 1.4) | 4.7 (SD 2.6) |
| Radial shock waves plus shared care | 34 | 6.9 (SD 1.9) | 4.9 (SD 2.9) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Between-arm difference at this timepoint; table groups ordered placebo then shockwave.
Unadjusted independent-samples t test. Missing data handled by the authors using ITT with last observation carried forward.
The reported between-group estimate is nonsignificant and its confidence interval includes zero; this does not establish equivalence.
Analysis details and source
Keep this between-group pointwise contrast distinct from within-arm changes reported in Tables 4 and 5. Fig. 1 reports 26 shockwave and 23 placebo participants followed up at week 4 (the Results text gives 20 of 70 lost by week 4); the analyses kept all 34 and 36 by carrying the last observation forward. Between-arm change-score comparison from Table 5: Placebo change 2.4 (SD 0.50) vs shockwave 1.9 (SD 0.44); difference 0.50 (95% CI -0.87 to 1.82), p=0.484. These change-score tests are separate from Table 3 pointwise comparisons. Within-arm Table 4 change from baseline (not a between-group effect): placebo 2.4 (0.50), 95% CI (1.4, 3.4), p=0.000; shockwave 1.9 (0.44), 95% CI (1.04, 2.8), p=0.000. Tables 4 and 5 print week 0 minus week 4, so positive values mean the pain score fell (improved). They head the bracketed values SD, but each printed 95% CI spans about twice that value either side of the mean, as expected of a standard error rather than an SD; values are kept as printed.
Numeric Pain Score (NPS)
Numeric Pain Score (NPS) points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Low-energy sham plus shared care | 36 | 7.1 (SD 1.4) | 4.5 (SD 2.6) |
| Radial shock waves plus shared care | 34 | 6.9 (SD 1.9) | 4.7 (SD 2.8) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Between-arm difference at this timepoint; table groups ordered placebo then shockwave.
Unadjusted independent-samples t test. Missing data handled by the authors using ITT with last observation carried forward.
The reported between-group estimate is nonsignificant and its confidence interval includes zero; this does not establish equivalence.
Analysis details and source
Keep this between-group pointwise contrast distinct from within-arm changes reported in Tables 4 and 5. Fig. 1 reports 18 shockwave and 12 placebo participants followed up at week 8; the analyses kept all 34 and 36 by carrying the last observation forward. Between-arm change-score comparison from Table 5: Placebo 2.6 (SD 0.49) vs shockwave 2.2 (SD 0.42); difference 0.40 (95% CI -0.89 to 1.73), p=0.525. These change-score tests are separate from Table 3 pointwise comparisons. Within-arm Table 4 change from baseline (not a between-group effect): placebo 2.6 (0.49), 95% CI (1.6, 3.6), p=0.000; shockwave 2.1 (0.42), 95% CI (1.3, 3.1), p=0.000. Table 4 reports active-arm change 2.1 (SD 0.42), while Table 5 reports 2.2 (SD 0.42); unresolved table discrepancy. Tables 4 and 5 print week 0 minus week 8, so positive values mean the pain score fell (improved). They head the bracketed values SD, but each printed 95% CI spans about twice that value either side of the mean, as expected of a standard error rather than an SD; values are kept as printed.
Numeric Pain Score (NPS)
Numeric Pain Score (NPS) points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Low-energy sham plus shared care | 36 | 7.1 (SD 1.4) | 3.5 (SD 2.6) |
| Radial shock waves plus shared care | 34 | 6.9 (SD 1.9) | 4.6 (SD 2.5) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Endpoint contrast is nonsignificant. Table 5 separately reports change from baseline: placebo 3.6 (SD 0.49) vs shockwave 2.3 (SD 0.40), difference 1.30 (95% CI 0.017 to 2.59; p=0.047), numerically favoring sham and conflicting with article conclusion.
Unadjusted independent-samples t test. Missing data handled by the authors using ITT with last observation carried forward.
No statistically significant pointwise between-group difference at this timepoint.
Analysis details and source
Keep this between-group pointwise contrast distinct from within-arm changes reported in Tables 4 and 5. Fig. 1 reports 15 shockwave and 18 placebo participants followed up at week 12; the analyses kept all 34 and 36 by carrying the last observation forward. Between-arm change-score comparison from Table 5: Placebo 3.6 (SD 0.49) vs shockwave 2.3 (SD 0.40); difference 1.30 (95% CI 0.017 to 2.59), p=0.047. These change-score tests are separate from Table 3 pointwise comparisons. Within-arm Table 4 change from baseline (not a between-group effect): placebo 3.6 (0.49), 95% CI (2.6, 4.6), p=0.000; shockwave 2.3 (0.40), 95% CI (1.5, 3.1), p=0.000. Tables 4 and 5 print week 0 minus week 12, so positive values mean the pain score fell (improved). They head the bracketed values SD, but each printed 95% CI spans about twice that value either side of the mean, as expected of a standard error rather than an SD; values are kept as printed.
Neck Disability Index (NDI)3 reported outcomes
Neck Disability Index (NDI)
Neck Disability Index (NDI) points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Low-energy sham plus shared care | 36 | 31.1 (SD 15.9) | 24.6 (SD 16.8) |
| Radial shock waves plus shared care | 34 | 29.6 (SD 15.5) | 24.1 (SD 14) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Between-arm difference at this timepoint; table groups ordered placebo then shockwave.
Unadjusted independent-samples t test. Missing data handled by the authors using ITT with last observation carried forward.
The reported between-group estimate is nonsignificant and its confidence interval includes zero; this does not establish equivalence.
Analysis details and source
Keep this between-group pointwise contrast distinct from within-arm changes reported in Tables 4 and 5. Fig. 1 reports 26 shockwave and 23 placebo participants followed up at week 4 (the Results text gives 20 of 70 lost by week 4); the analyses kept all 34 and 36 by carrying the last observation forward. Between-arm change-score comparison from Table 5: Placebo change 6.47 (SD 2.6) vs shockwave 5.56 (SD 2.1); difference 0.90 (95% CI -5.87 to 7.6), p=0.791. These change-score tests are separate from Table 3 pointwise comparisons. Within-arm Table 4 change from baseline (not a between-group effect): placebo 6.47 (2.6), 95% CI (1.15, 11.78), p=0.010; shockwave 5.56 (2.1), 95% CI (1.23, 9.90), p=0.010. Tables 4 and 5 print week 0 minus week 4, so positive values mean the disability score fell (improved). They head the bracketed values SD, but each printed 95% CI spans about twice that value either side of the mean, as expected of a standard error rather than an SD; values are kept as printed.
Neck Disability Index (NDI)
Neck Disability Index (NDI) points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Low-energy sham plus shared care | 36 | 31.1 (SD 15.9) | 24.9 (SD 17) |
| Radial shock waves plus shared care | 34 | 29.6 (SD 15.5) | 24.9 (SD 16.6) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Between-arm difference at this timepoint; table groups ordered placebo then shockwave.
Unadjusted independent-samples t test. Missing data handled by the authors using ITT with last observation carried forward.
The reported between-group estimate is nonsignificant and its confidence interval includes zero; this does not establish equivalence.
Analysis details and source
Keep this between-group pointwise contrast distinct from within-arm changes reported in Tables 4 and 5. Fig. 1 reports 18 shockwave and 12 placebo participants followed up at week 8; the analyses kept all 34 and 36 by carrying the last observation forward. Between-arm change-score comparison from Table 5: Placebo 6.13 (SD 2.7) vs shockwave 4.75 (SD 2.6); difference 1.38 (95% CI -6.2 to 8.96), p=0.718. These change-score tests are separate from Table 3 pointwise comparisons. Within-arm Table 4 change from baseline (not a between-group effect): placebo 6.13 (2.7), 95% CI (0.56, 11.72), p=0.030; shockwave 4.75 (2.6), 95% CI (− 0.56, 10.07), p=0.070. Tables 4 and 5 print week 0 minus week 8, so positive values mean the disability score fell (improved). They head the bracketed values SD, but each printed 95% CI spans about twice that value either side of the mean, as expected of a standard error rather than an SD; values are kept as printed. The Results text gives p=0.005 for the placebo week 0–8 change, where Table 4 prints p=0.030.
Neck Disability Index (NDI)
Neck Disability Index (NDI) points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Low-energy sham plus shared care | 36 | 31.1 (SD 15.9) | 22.8 (SD 18) |
| Radial shock waves plus shared care | 34 | 29.6 (SD 15.5) | 24.5 (SD 15.7) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Between-arm difference at this timepoint; table groups ordered placebo then shockwave.
Unadjusted independent-samples t test. Missing data handled by the authors using ITT with last observation carried forward.
The reported between-group estimate is nonsignificant and its confidence interval includes zero; this does not establish equivalence.
Analysis details and source
Keep this between-group pointwise contrast distinct from within-arm changes reported in Tables 4 and 5. Fig. 1 reports 15 shockwave and 18 placebo participants followed up at week 12; the analyses kept all 34 and 36 by carrying the last observation forward. Between-arm change-score comparison from Table 5: Placebo 8.22 (SD 2.9) vs shockwave 5.12 (SD 2.7); difference 3.09 (95% CI -5.01 to 11.20), p=0.449. These change-score tests are separate from Table 3 pointwise comparisons. Within-arm Table 4 change from baseline (not a between-group effect): placebo 8.22 (2.9), 95% CI (2.15, 14.28), p=0.000; shockwave 5.12 (2.7), 95% CI (-.44, 10.69), p=0.060. Tables 4 and 5 print week 0 minus week 12, so positive values mean the disability score fell (improved). They head the bracketed values SD, but each printed 95% CI spans about twice that value either side of the mean, as expected of a standard error rather than an SD; values are kept as printed.
Pressure pain threshold3 reported outcomes
Pressure pain threshold
Pressure pain threshold kg/cm²
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Low-energy sham plus shared care | 36 | 2.8 (SD 1.7) | 3.3 (SD 2.1) |
| Radial shock waves plus shared care | 34 | 2.4 (SD 1.4) | 3 (SD 1.3) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Between-arm difference at this timepoint; table groups ordered placebo then shockwave.
Unadjusted independent-samples t test. Missing data handled by the authors using ITT with last observation carried forward.
The reported between-group estimate is nonsignificant and its confidence interval includes zero; this does not establish equivalence.
Analysis details and source
Keep this between-group pointwise contrast distinct from within-arm changes reported in Tables 4 and 5. Fig. 1 reports 26 shockwave and 23 placebo participants followed up at week 4 (the Results text gives 20 of 70 lost by week 4); the analyses kept all 34 and 36 by carrying the last observation forward. Between-arm change-score comparison from Table 5: Placebo change -0.48 (SD 0.21) vs shockwave -0.5 (SD 0.17); difference 0.10 (95% CI -0.46 to 0.66), p=0.718. These change-score tests are separate from Table 3 pointwise comparisons. Within-arm Table 4 change from baseline (not a between-group effect): placebo − 0.48 (0.21), 95% CI (− 0.93, − 0.04), p=0.030; shockwave − 0.5 (0.17), 95% CI (− 0.94, − 0.23), p=0.002. Tables 4 and 5 print week 0 minus week 4, so negative values mean the pressure pain threshold rose (improved). They head the bracketed values SD, but each printed 95% CI spans about twice that value either side of the mean, as expected of a standard error rather than an SD; values are kept as printed. Table 5 prints a difference of 0.10, although its own group values give about 0.02 (−0.48 minus −0.5); kept as printed.
Pressure pain threshold
Pressure pain threshold kg/cm²
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Low-energy sham plus shared care | 36 | 2.8 (SD 1.7) | 3.3 (SD 2.1) |
| Radial shock waves plus shared care | 34 | 2.4 (SD 1.4) | 3 (SD 1.4) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Between-arm difference at this timepoint; table groups ordered placebo then shockwave.
Unadjusted independent-samples t test. Missing data handled by the authors using ITT with last observation carried forward.
The reported between-group estimate is nonsignificant and its confidence interval includes zero; this does not establish equivalence.
Analysis details and source
Keep this between-group pointwise contrast distinct from within-arm changes reported in Tables 4 and 5. Fig. 1 reports 18 shockwave and 12 placebo participants followed up at week 8; the analyses kept all 34 and 36 by carrying the last observation forward. Between-arm change-score comparison from Table 5: Placebo -0.51 (SD 0.21) vs shockwave -0.5 (SD 0.17); difference 0.02 (95% CI -0.55 to 0.58), p=0.951. These change-score tests are separate from Table 3 pointwise comparisons. Within-arm Table 4 change from baseline (not a between-group effect): placebo − 0.51 (0.21), 95% CI (− 0.96, − 0.08), p=0.020; shockwave − 0.5 (0.17), 95% CI (− 0.90, − 0.17), p=0.005. Tables 4 and 5 print week 0 minus week 8, so negative values mean the pressure pain threshold rose (improved). They head the bracketed values SD, but each printed 95% CI spans about twice that value either side of the mean, as expected of a standard error rather than an SD; values are kept as printed.
Pressure pain threshold
Pressure pain threshold kg/cm²
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Low-energy sham plus shared care | 36 | 2.8 (SD 1.7) | 3.4 (SD 2.1) |
| Radial shock waves plus shared care | 34 | 2.4 (SD 1.4) | 3.1 (SD 1.5) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Between-arm difference at this timepoint; table groups ordered placebo then shockwave.
Unadjusted independent-samples t test. Missing data handled by the authors using ITT with last observation carried forward.
The reported between-group estimate is nonsignificant and its confidence interval includes zero; this does not establish equivalence.
Analysis details and source
Keep this between-group pointwise contrast distinct from within-arm changes reported in Tables 4 and 5. Fig. 1 reports 15 shockwave and 18 placebo participants followed up at week 12; the analyses kept all 34 and 36 by carrying the last observation forward. Between-arm change-score comparison from Table 5: Placebo -0.66 (SD 0.25) vs shockwave -0.6 (SD 0.17); difference 0.02 (95% CI -0.59 to 0.64), p=0.942. These change-score tests are separate from Table 3 pointwise comparisons. Within-arm Table 4 change from baseline (not a between-group effect): placebo − 0.66 (0.25), 95% CI (− 1.17, − 0.15), p=0.010; shockwave − 0.6 (0.17), 95% CI (− 1.04, − 0.32), p=0.000. Tables 4 and 5 print week 0 minus week 12, so negative values mean the pressure pain threshold rose (improved). They head the bracketed values SD, but each printed 95% CI spans about twice that value either side of the mean, as expected of a standard error rather than an SD; values are kept as printed. Table 5 prints a difference of 0.02, although its own group values give about −0.06 (−0.66 minus −0.6); kept as printed.
SF-12 Physical Component Score3 reported outcomes
SF-12 Physical Component Score
SF-12 Physical Component Score score points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Low-energy sham plus shared care | 36 | 40.9 (SD 9) | 42.7 (SD 10.8) |
| Radial shock waves plus shared care | 34 | 40.8 (SD 8.9) | 41.8 (SD 7.5) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Between-arm difference at this timepoint; table groups ordered placebo then shockwave.
Unadjusted independent-samples t test. Missing data handled by the authors using ITT with last observation carried forward.
The reported between-group estimate is nonsignificant and its confidence interval includes zero; this does not establish equivalence.
Analysis details and source
Keep this between-group pointwise contrast distinct from within-arm changes reported in Tables 4 and 5. Fig. 1 reports 26 shockwave and 23 placebo participants followed up at week 4 (the Results text gives 20 of 70 lost by week 4); the analyses kept all 34 and 36 by carrying the last observation forward. Between-arm change-score comparison from Table 5: Placebo change -1.8 (SD 1.1) vs shockwave -0.9 (SD 1.1); difference -0.91 (95% CI -4.1 to 2.31), p=0.5753. These change-score tests are separate from Table 3 pointwise comparisons. Within-arm Table 4 change from baseline (not a between-group effect): placebo − 1.8 (1.1), 95% CI (− 4.24, 0.50), p=0.110; shockwave − 0.9 (1.1), 95% CI (− 3.22, 1.29), p=0.392. Tables 4 and 5 print week 0 minus week 4, so negative values mean the score rose (improved). They head the bracketed values SD, but each printed 95% CI spans about twice that value either side of the mean, as expected of a standard error rather than an SD; values are kept as printed.
SF-12 Physical Component Score
SF-12 Physical Component Score score points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Low-energy sham plus shared care | 36 | 40.9 (SD 9) | 43.7 (SD 10.8) |
| Radial shock waves plus shared care | 34 | 40.8 (SD 8.9) | 40.7 (SD 9.1) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Between-arm difference at this timepoint; table groups ordered placebo then shockwave.
Unadjusted independent-samples t test. Missing data handled by the authors using ITT with last observation carried forward.
The reported between-group estimate is nonsignificant and its confidence interval includes zero; this does not establish equivalence.
Analysis details and source
Keep this between-group pointwise contrast distinct from within-arm changes reported in Tables 4 and 5. Fig. 1 reports 18 shockwave and 12 placebo participants followed up at week 8; the analyses kept all 34 and 36 by carrying the last observation forward. Between-arm change-score comparison from Table 5: Placebo -2.8 (SD 1.1) vs shockwave 0.07 (SD 1.4); difference -2.9 (95% CI -6.6 to 0.84); p-value not shown in Table 5. These change-score tests are separate from Table 3 pointwise comparisons. Within-arm Table 4 change from baseline (not a between-group effect): placebo − 2.8 (1.1), 95% CI (− 5.11, − 0.51), p=0.010; shockwave 0.07 (1.4), 95% CI (− 2.97, 3.12), p=0.959. Tables 4 and 5 print week 0 minus week 8, so negative values mean the score rose (improved). They head the bracketed values SD, but each printed 95% CI spans about twice that value either side of the mean, as expected of a standard error rather than an SD; values are kept as printed.
SF-12 Physical Component Score
SF-12 Physical Component Score score points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Low-energy sham plus shared care | 36 | 40.9 (SD 9) | 44.3 (SD 9.3) |
| Radial shock waves plus shared care | 34 | 40.8 (SD 8.9) | 42.3 (SD 9.9) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Between-arm difference at this timepoint; table groups ordered placebo then shockwave.
Unadjusted independent-samples t test. Missing data handled by the authors using ITT with last observation carried forward.
The reported between-group estimate is nonsignificant and its confidence interval includes zero; this does not establish equivalence.
Analysis details and source
Keep this between-group pointwise contrast distinct from within-arm changes reported in Tables 4 and 5. Fig. 1 reports 15 shockwave and 18 placebo participants followed up at week 12; the analyses kept all 34 and 36 by carrying the last observation forward. Between-arm change-score comparison from Table 5: Placebo -3.4 (SD 1.4) vs shockwave -1.40 (SD 1.7); difference -1.93 (95% CI -6.51 to 2.64), p=0.4020. These change-score tests are separate from Table 3 pointwise comparisons. Within-arm Table 4 change from baseline (not a between-group effect): placebo − 3.4 (1.4), 95% CI (− 6.38, − 0.45), p=0.020; shockwave − 1.4 (1.7), 95% CI (− 5.11, 2.15), p=0.412. Tables 4 and 5 print week 0 minus week 12, so negative values mean the score rose (improved). They head the bracketed values SD, but each printed 95% CI spans about twice that value either side of the mean, as expected of a standard error rather than an SD; values are kept as printed.
SF-12 Mental Component Score3 reported outcomes
SF-12 Mental Component Score
SF-12 Mental Component Score score points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Low-energy sham plus shared care | 36 | 45.7 (SD 12.2) | 47.1 (SD 10.9) |
| Radial shock waves plus shared care | 34 | 45 (SD 10.6) | 46.4 (SD 10.9) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Between-arm difference at this timepoint; table groups ordered placebo then shockwave.
Unadjusted independent-samples t test. Missing data handled by the authors using ITT with last observation carried forward.
The reported between-group estimate is nonsignificant and its confidence interval includes zero; this does not establish equivalence.
Analysis details and source
Keep this between-group pointwise contrast distinct from within-arm changes reported in Tables 4 and 5. Fig. 1 reports 26 shockwave and 23 placebo participants followed up at week 4 (the Results text gives 20 of 70 lost by week 4); the analyses kept all 34 and 36 by carrying the last observation forward. Between-arm change-score comparison from Table 5: Placebo change -1.4 (SD 6.3) vs shockwave -1.42 (SD 8.2); difference 0.01 (95% CI -3.48 to 3.49), p=0.9978. These change-score tests are separate from Table 3 pointwise comparisons. Within-arm Table 4 change from baseline (not a between-group effect): placebo − 1.4 (6.3), 95% CI (− 3.55, 0.70), p=0.183; shockwave − 1.42 (8.2), 95% CI (− 4.30, 1.45), p=0.320. Tables 4 and 5 print week 0 minus week 4, so negative values mean the score rose (improved).
SF-12 Mental Component Score
SF-12 Mental Component Score score points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Low-energy sham plus shared care | 36 | 45.7 (SD 12.2) | 46.3 (SD 11.8) |
| Radial shock waves plus shared care | 34 | 45 (SD 10.6) | 45.3 (SD 9.9) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Between-arm difference at this timepoint; table groups ordered placebo then shockwave.
Unadjusted independent-samples t test. Missing data handled by the authors using ITT with last observation carried forward.
The reported between-group estimate is nonsignificant and its confidence interval includes zero; this does not establish equivalence.
Analysis details and source
Keep this between-group pointwise contrast distinct from within-arm changes reported in Tables 4 and 5. Fig. 1 reports 18 shockwave and 12 placebo participants followed up at week 8; the analyses kept all 34 and 36 by carrying the last observation forward. Between-arm change-score comparison from Table 5: Placebo -0.6 (SD 7.5) vs shockwave -0.30 (SD 10.4); difference -0.25 (95% CI -4.58 to 4.09), p=0.9103. These change-score tests are separate from Table 3 pointwise comparisons. Within-arm Table 4 change from baseline (not a between-group effect): placebo − 0.6 (7.5), 95% CI (− 3.17, 1.91), p=0.617; shockwave − 0.3 (10.4), 95% CI (− 4.05, 3.28), p=0.832. Tables 4 and 5 print week 0 minus week 8, so negative values mean the score rose (improved).
SF-12 Mental Component Score
SF-12 Mental Component Score score points
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Low-energy sham plus shared care | 36 | 45.7 (SD 12.2) | 48.4 (SD 10.7) |
| Radial shock waves plus shared care | 34 | 45 (SD 10.6) | 47.5 (SD 10.5) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Between-arm difference at this timepoint; table groups ordered placebo then shockwave.
Unadjusted independent-samples t test. Missing data handled by the authors using ITT with last observation carried forward.
The reported between-group estimate is nonsignificant and its confidence interval includes zero; this does not establish equivalence.
Analysis details and source
Keep this between-group pointwise contrast distinct from within-arm changes reported in Tables 4 and 5. Fig. 1 reports 15 shockwave and 18 placebo participants followed up at week 12; the analyses kept all 34 and 36 by carrying the last observation forward. Between-arm change-score comparison from Table 5: Placebo -2.7 (SD 8.8) vs shockwave -2.50 (SD 9.7); difference -0.18 (95% CI -4.62 to 4.26), p=0.9354. These change-score tests are separate from Table 3 pointwise comparisons. Within-arm Table 4 change from baseline (not a between-group effect): placebo − 2.7 (8.8), 95% CI (− 5.74, 0.21), p=0.06; shockwave − 2.5 (9.7), 95% CI (− 5.99,.83), p=0.133. Tables 4 and 5 print week 0 minus week 12, so negative values mean the score rose (improved).
How treatment was delivered
Storz Medical radial shock-wave device.
Methods report 1.5 bar and 0.57 mJ/mm²; abstract reports 1.5 bar and 0.068 mJ/mm². Unresolved; no conversion applied.
0.3 bar (0.01 mJ/mm²), described as nontherapeutic no-energy shock.
2,000 pulses/session (stated only in the Abstract, the same Abstract whose energy figure conflicts with the Methods), 15 Hz, D20 transmitter, 20-mm headpiece.
Six weekly sessions, directly to identified trapezius/periscapular trigger or painful points.
Education, stretches, exercises and home-exercise guidance in both arms.
Not reported.
The Results section says one participant discontinued treatment at week 4 because pain increased. The adverse-events section says two active-arm participants had temporary pain sensitivity, neither dropped out, and no adverse-effect dropouts occurred. The report does not clarify whether treatment discontinuation differed from study dropout. The aftercare paragraph says: "Only one participant reported increased pain after the intervention, and no other serious adverse effects were reported in this study."
The authors planned 120 participants but enrolled 70; 20/70 (29%) were lost by week 4, and last observation carried forward was used.
Table 5 reports a week-12 NPS change contrast at p=0.047 favoring sham while narrative/conclusion say no between-group difference. Active energy differs between abstract and Methods.
Supplementary Material 1 was cited but unavailable.
The participant flow is internally inconsistent: Results reports 102 approached and 32 excluded (70 randomized), whereas Fig. 1 reports 82 assessed and 10 excluded although its listed reasons sum to 12. At week 4, Fig. 1 shows 26 active plus 23 sham followed (21 not followed), versus Results’ 20/70 lost; later sham counts also rise from 12 at week 8 to 18 at week 12. The figure ends with 34/36 analyzed. The available text does not resolve these discrepancies.
Findings checked against the full source by a separate automated reviewer.
Limitations noted by the reviewers
- The authors planned 120 participants but enrolled 70; 20/70 (29%) were lost by week 4, and last observation carried forward was used.
- Table 5 reports a week-12 NPS change contrast at p=0.047 favoring sham while narrative/conclusion say no between-group difference. Active energy differs between abstract and Methods.
- Supplementary Material 1 was cited but unavailable.
- The participant flow is internally inconsistent: Results reports 102 approached and 32 excluded (70 randomized), whereas Fig. 1 reports 82 assessed and 10 excluded although its listed reasons sum to 12. At week 4, Fig. 1 shows 26 active plus 23 sham followed (21 not followed), versus Results’ 20/70 lost; later sham counts also rise from 12 at week 8 to 18 at week 12. The figure ends with 34/36 analyzed. The available text does not resolve these discrepancies.
Source-linked critical reading
Randomized trial with sham / placebo control
Technical source record
Complete full-text XML, all article sections, Tables 1–5, and the Fig. 1 caption read; Figure 1 was visually inspected and compared with the Results narrative. Supplementary Material 1 was cited but unavailable.
Both groups improved on some within-arm measures, but that does not show an added shock-wave benefit. Most between-group comparisons were nonsignificant; one week-12 pain-change table result favors sham at p=0.047, contrary to the article’s broad conclusion. The added effect remains uncertain.
In adults with neck or upper-back myofascial pain, does radial shock-wave therapy added to shared physical therapy and exercise improve pain, disability, pressure pain threshold, or quality of life more than low-energy sham?
Prospectively registered, two-arm, patient- and assessor-blinded randomized placebo-controlled trial in one Riyadh rehabilitation department. Computer sequence used blocks of four and opaque sealed envelopes; blinded staff collected outcomes and a blinded statistician analyzed results. Seventy participants were randomized, 34 to active treatment and 36 to sham.
Adults aged at least 19 had neck and/or upper-back pain with palpable tenderness and active trigger points, confirmed by a treating physical therapist. The trial excluded multiple conditions and treatments, including recent trigger-point steroid injection, fibromyalgia, radiculopathy/myelopathy, pregnancy, and recent neck/shoulder surgery. Mean ages were about 45 years; symptom duration averaged roughly 47–51 weeks.
The active arm received six radial shock-wave sessions one week apart with a Storz device, 2,000 pulses at 15 Hz, using a 20-mm D20 transmitter over identified trapezius/periscapular trigger or painful points. Methods report 1.5 bar and 0.57 mJ/mm², while the abstract reports 1.5 bar and 0.068 mJ/mm². Both groups also received the same stretches, exercises and home program.
The sham group received the same six-session schedule and shared physical therapy, but the device delivered a no-energy shock at 0.3 bar (0.01 mJ/mm²); participants heard the machine. This estimates the incremental effect of a higher-energy radial dose on top of exercise and therapist contact, not ESWT versus no treatment. The abstract/methods report inconsistent active energy values (0.068 versus 0.57 mJ/mm²), which requires source clarification.
Primary outcomes were Numeric Pain Score (NPS, 0–10) and Neck Disability Index (NDI, 0–50), assessed at baseline and weeks 4, 8, and 12. Secondary outcomes were pressure pain threshold (PPT) and SF-12 physical and mental scores. The study defined NPS MCID as 2 points or 30%; NDI MCID as 3.5–5 points.
At week 12, NPS means were 3.5 (SD 2.6) with sham and 4.6 (SD 2.5) with radial shock waves; the endpoint contrast was -1.10 (95% CI -2.33 to 0.13; p=0.078). Table 5 separately reports change from baseline of 3.6 (SD 0.49) with sham versus 2.3 (SD 0.40) with active treatment, difference 1.30 (95% CI 0.017 to 2.59; p=0.047), numerically favoring sham and conflicting with the broader no-difference conclusion. At week 12, between-group endpoint contrasts were nonsignificant for NDI (-1.68; 95% CI -9.77 to 6.40; p=0.679), PPT (0.314 kg/cm²; 95% CI -0.57 to 1.2; p=0.479), SF-12 physical (1.9; 95% CI -2.57 to 6.55; p=0.387) and mental (0.89; 95% CI -4.20 to 5.90; p=0.726). Table 5 does not report a p-value for SF-12 physical change at weeks 0–8. Within-arm changes do not establish superiority.
The report says one participant discontinued treatment at week 4 because pain increased. A later adverse-events paragraph says two active-arm participants had temporary pain sensitivity and neither dropped out; it also says there were no adverse-effect dropouts. The report does not clarify whether “discontinued treatment” differs from study dropout.
The trial recruited 70 although its calculation called for 120 to allow for attrition and provide 90% power; 20/70 (29%) were lost by week 4. Missing outcomes were handled with last observation carried forward, which can bias estimates. Most importantly, the report’s conclusion of no between-group difference conflicts with Table 5’s borderline 12-week pain-change result favoring sham; and active energy density differs between the abstract and Methods by nearly an order of magnitude (0.068 vs 0.57 mJ/mm²). The report’s “no adverse-effect dropout” statement also conflicts with the pain-related treatment discontinuation. The participant flow is internally inconsistent: Results reports 102 approached and 32 excluded (70 randomized), whereas Fig. 1 reports 82 assessed and 10 excluded although its listed reasons sum to 12. At week 4, Fig. 1 shows 26 active plus 23 sham followed (21 not followed), versus Results’ 20/70 lost; later sham counts also rise from 12 at week 8 to 18 at week 12. The figure ends with 34/36 analyzed. The available text does not resolve these discrepancies.
This pragmatic comparison applies to adults with diagnosed neck/upper-back MPS receiving a substantial shared exercise and therapist-contact program. It does not show that radial shock waves alone are effective. The comparator was a low-energy sham, while the active-dose discrepancy and outcome-reporting conflict weaken confidence in the exact treatment contrast.
A transparent replication should reconcile device energy reporting and prespecify whether primary inference uses endpoint or change-score comparisons. It should recruit the planned sample, reduce and characterize follow-up losses, use robust missing-data methods, report blinded between-group estimates and confidence intervals, clarify the pain-related discontinuation, and separate effects of shock waves from shared exercise/contact.
Read the original article.
The effectiveness of radial shockwave therapy on myofascial pain syndrome: a two-armed, randomized double-blind placebo-controlled trial.
Ogbeivor C, AlMubarak H, Akomolafe T, Alkahtani H, AlMugizel H, Marin I, Aldosari H, Aldhwayan N, Mohamed G, Alobthani K. · BMC musculoskeletal disorders · 2025
Open the publication record or available full-text files.
Checked against the full paper · Clinical review: Dr. Marcus Yu Bin Pai, · How we review →