Chronic plantar fasciitis: success rates pooled by shockwave type · Review · 2017
Does shockwave help long-lasting heel pain more than sham?
In this review of 9 sham-controlled trials, more people reached each trial’s success goal after shockwave, over 6 weeks to 6 months, most consistently with focused devices. One trial drove the size of the radial result; without it, radial still favoured shockwave.
Helped more than sham- trials against sham
- 9
- adults with long-lasting heel pain
- 935
- focused-device trials; 3 classed as radial
- 6
Good to know: Trials that gave everyone a numbing injection were left out.
Sun J, Gao F, Wang Y, et al. Medicine (Baltimore), 2017 · Read the paper ↗
How this study worked
What happened
- studies
- 9
- people
- 935
- Compared with
- Sham shockwave
Chance of success vs sham
Odds ratio · above 1 means more successes with shockwave
Show the numbers
| Odds ratio | likely range | |
|---|---|---|
| All 9 trials | 2.58 | 1.97 to 3.39 |
| 6 focused trials | 2.17 | 1.49 to 3.16 |
| Radial, without the outlying trial | 2.39 | 1.57 to 3.65 |
What the trials used
- Focused regimens
- Rompe 1996 0.06 mJ/mm², 1,000 pulses ×3; Speed 2003 0.12, 1,500 ×3; Rompe 2003 0.16, 2,100 ×3; Marks 2008 0.16, 1,500 ×3; Gollwitzer 2007 and 2015 0.25, 2,000 ×3. Trials applying local anaesthesia to all patients were excluded.
- Regimens classed as radial
- Gerdesmeyer 2008 0.16 mJ/mm², 2,000 pulses ×3; Ibrahim 2010 0.16, 2,000 ×2; Malay 2006 a total of 3,800 pulses, intensity not available.
- Sham methods
- Sham treatments: air-filled or foil standoffs, foam-insulated membrane, sound-reflecting pad, zero- or minimal-energy pulses, a heel clasp, a placebo handpiece, or the applicator held 1 cm away.
- Device classification check
- Malay 2006 describes an electrohydraulic spark-gap device with a semi-ellipsoid reflector; the trial report does not call it radial, although it says the energy is dispersed over a treatment area large enough that no targeting is needed. Marks 2008 is listed as focused although its reference title describes a new-generation pneumatic device.
Who took part
Adults with long-lasting plantar heel pain, with average ages of 40 to 59, in 9 trials from 1996 to 2015.
Side effects
No serious problems were recorded; a few people had discomfort, pain, swelling or bruising.
All the numbers and the full appraisal
The results, in numbers.
Adults with chronic plantar fasciitis in 9 sham-controlled RCTs (1996–2015), 935 patients: 6 trials of focused and 3 classed as radial shockwaves. Trials giving local anaesthesia to all patients were excluded. Group mean ages 40 to 58.9 years; follow-up 6 weeks to 6 months.
Treatment success, all shockwaves4 reported outcomes
Treatment success, all shockwaves
Treatment success: >50% or >60% VAS improvement from baseline or similar, per trial odds of treatment success
Shockwaves versus sham
9 studies · 935 participants in this estimate · I²=38%; Chi²=12.81, df=8, P=0.12
Mantel–Haenszel fixed-effect model
Logarithmic ratio scale.
Across 9 trials more patients reached the trial-defined success threshold after shockwaves than after sham (275/506 versus 141/429 in Figure 4).
Analysis details and source
Success definitions differed (>50%, >60%, VAS <4 cm; Rompe 1996 had no stated criterion). Figure 4 uses 12/22 versus 6/23 for Rompe 2003 and 17/46 versus 10/42 for Speed 2003, whereas Table 1 lists 12/19, 6/20, 17/42 and 10/34. Ibrahim 2010 (25/25 versus 4/25) has OR 243.67.
Treatment success, focused shockwaves
Treatment success: >50% or >60% VAS improvement from baseline or similar, per trial odds of treatment success
Shockwaves versus sham
6 studies · 474 participants in this estimate · I²=0%; Chi²=2.22, df=5, P=0.82
Mantel–Haenszel fixed-effect model
Logarithmic ratio scale.
In 6 focused-device trials more patients reached success after shockwaves (127/244 versus 77/230), with consistent results across trials. Speed 2003 and Marks 2008 alone showed no significant difference.
Analysis details and source
The text describes 474 patients “with 244 events”; Figure 5 shows 244 as the treated-arm total and 127 events. Marks 2008 is classed as focused although its title describes a pneumatic device.
Treatment success, radial shockwaves
Treatment success: >50% or >60% VAS improvement from baseline or similar, per trial odds of treatment success
Shockwaves versus sham
3 studies · 461 participants in this estimate · I²=81%; Tau²=0.86; Chi²=10.47, df=2, P=0.005
Random-effects model
Logarithmic ratio scale.
The pooled radial estimate favoured shockwaves but with a very wide interval and large between-trial differences.
Analysis details and source
Three trials: Gerdesmeyer 2008 OR 2.11 (1.27–3.52), Ibrahim 2010 OR 243.67 (12.41–4785.60) and Malay 2006 OR 3.07. Malay 2006 describes an electrohydraulic semi-ellipsoid device, which its report does not call radial. Text gives 461 patients “with 262 events”; Figure 6 shows 262 as treated total and 148 events.
Radial success without Ibrahim 2010
Treatment success: >50% or >60% VAS improvement from baseline or similar, per trial odds of treatment success
Radial shockwaves versus sham, Ibrahim 2010 excluded
2 studies
Sensitivity analysis; model not stated
Logarithmic ratio scale.
Removing the outlying Ibrahim trial left Gerdesmeyer 2008 and Malay 2006, still favouring shockwaves with a narrower interval.
Analysis details and source
Sensitivity analysis reported only in the Discussion; model and participant count not stated. One of the two remaining trials (Malay) may not be a radial device. Excluding Malay instead gave OR 18.34 (0.14–2413.75), I²=90%, which the authors still describe as efficacy.
Reduction in pain scale, all shockwaves2 reported outcomes
Reduction in pain scale, all shockwaves
Reduction in pain scale (VAS-based, trial-specific) standardized mean difference units
Shockwaves versus sham
4 studies · I²=96%; Tau²=1.00; Chi²=72.62, df=3, P<0.00001
Random-effects model
Pooled pain reduction was larger with shockwaves on average, but the 95% CI reaches −0.01 and the trials differed greatly (I²=96%).
Analysis details and source
Positive SMD means greater pain reduction with shockwaves, although the axis labels this side “favours control”, and the abstract calls it a “markedly lower” SMD. The text says the trials reported standard errors; Gollwitzer 2015 values (6.8, 7.2) entered in the SD column match Table 1 values in parentheses. If these are SEs, its SMD of 2.02 would be overstated (extractor interpretation). Figures 7 and 8 give Gollwitzer 125 and 124 treated.
Reduction in pain scale, focused shockwaves
Reduction in pain scale (VAS-based, trial-specific) standardized mean difference units
Focused shockwaves versus sham
3 studies · I²=87%; Tau²=0.53; Chi²=14.87, df=2, P=0.0006
Random-effects model
In 3 focused-device trials pain reduction was larger with shockwaves, with very high between-trial differences.
Analysis details and source
Driven by Gollwitzer 2015 (SMD 2.02); Marks 2008 alone 0.76 (−0.09 to 1.61). Excluding Gollwitzer from the 4-trial analysis gave SMD 0.57 (0.12–1.01), p=.01 (Discussion).
How treatment was delivered
Rompe 1996 0.06 mJ/mm², 1,000 pulses ×3; Speed 2003 0.12, 1,500 ×3; Rompe 2003 0.16, 2,100 ×3; Marks 2008 0.16, 1,500 ×3; Gollwitzer 2007 and 2015 0.25, 2,000 ×3. Trials applying local anaesthesia to all patients were excluded.
Gerdesmeyer 2008 0.16 mJ/mm², 2,000 pulses ×3; Ibrahim 2010 0.16, 2,000 ×2; Malay 2006 a total of 3,800 pulses, intensity not available.
Sham treatments: air-filled or foil standoffs, foam-insulated membrane, sound-reflecting pad, zero- or minimal-energy pulses, a heel clasp, a placebo handpiece, or the applicator held 1 cm away.
Malay 2006 describes an electrohydraulic spark-gap device with a semi-ellipsoid reflector; the trial report does not call it radial, although it says the energy is dispersed over a treatment area large enough that no targeting is needed. Marks 2008 is listed as focused although its reference title describes a new-generation pneumatic device.
Seven of 9 trials described complications. No serious adverse events (such as haematoma or infection) were recorded; a few patients had discomfort, pain, swelling or bruising during or after treatment. No pooled harm estimate was made, and two trials did not describe complications, so the authors’ statement that ESWT is safe rests on limited reporting.
Title versus conclusion: the title says ESWT “is effective”, but the abstract and conclusion say no firm conclusion can be drawn about ESWT overall or radial therapy. The Discussion’s opening says ESWT and radial therapy “might be more effective than placebo and FSW therapy”, while the conclusion favours focused therapy.
Reporting errors: subgroup “events” in the text (244, 262) are treated-arm totals in Figures 5–6; Table 1 and Figure 4 disagree on Rompe 2003 and Speed 2003 counts; the abstract calls a positive SMD “markedly lower”; forest-plot axis labels are reversed relative to the data.
Pain-scale synthesis: only 4 trials, I²=96%, mixed units in Table 1 (percentages and points), and values described as standard errors appear in the SD column (extractor interpretation).
Device classification: Malay 2006 (electrohydraulic) is analysed as radial and Marks 2008 (pneumatic per its title) as focused; Ibrahim 2010 (25/25 successes) dominates radial heterogeneity.
Author-reported limits: English-only search to July 2016; some participants received local anaesthesia for intolerable pain; functional recovery could not be analysed; no direct focused-versus-radial trials pooled. Risk of bias was mostly rated low; Rompe 1996 was unclear in four domains.
Funding: National Natural Science Foundation of China, Beijing Natural Science Foundation and China-Japan Friendship Hospital research funds; the authors report no conflicts of interest.
Findings checked against the full source by a separate automated reviewer.
Limitations noted by the reviewers
- Title versus conclusion: the title says ESWT “is effective”, but the abstract and conclusion say no firm conclusion can be drawn about ESWT overall or radial therapy. The Discussion’s opening says ESWT and radial therapy “might be more effective than placebo and FSW therapy”, while the conclusion favours focused therapy.
- Reporting errors: subgroup “events” in the text (244, 262) are treated-arm totals in Figures 5–6; Table 1 and Figure 4 disagree on Rompe 2003 and Speed 2003 counts; the abstract calls a positive SMD “markedly lower”; forest-plot axis labels are reversed relative to the data.
- Pain-scale synthesis: only 4 trials, I²=96%, mixed units in Table 1 (percentages and points), and values described as standard errors appear in the SD column (extractor interpretation).
- Device classification: Malay 2006 (electrohydraulic) is analysed as radial and Marks 2008 (pneumatic per its title) as focused; Ibrahim 2010 (25/25 successes) dominates radial heterogeneity.
- Author-reported limits: English-only search to July 2016; some participants received local anaesthesia for intolerable pain; functional recovery could not be analysed; no direct focused-versus-radial trials pooled. Risk of bias was mostly rated low; Rompe 1996 was unclear in four domains.
- Funding: National Natural Science Foundation of China, Beijing Natural Science Foundation and China-Japan Friendship Hospital research funds; the authors report no conflicts of interest.
Source-linked critical reading
Systematic review and meta-analysis
Technical source record
Complete local PDF read (all 7 pages), including Table 1, all eight figures checked on page images, discussion, funding and references. The included Malay 2006 trial was also read in full to check its device classification; other included trials were not reread.
In sham-controlled trials of long-standing plantar heel pain, more people reached each trial’s success threshold after shockwaves, most consistently with focused devices, over 6 weeks to 6 months. The size of the radial result comes largely from one small trial; without it radial treatment still came out ahead of sham, though by less, and the larger radial trial on its own also favoured shockwaves; and errors in the paper mean the exact numbers should be read with care.
Compared with sham, do shockwaves in general, and focused or radial shockwaves separately, increase treatment success and pain reduction in chronic plantar fasciitis?
Systematic review and meta-analysis following PRISMA: PubMed, Medline, Embase, Web of Science and Cochrane Library to July 2016, English only. Nine sham-controlled RCTs (935 patients) included; Cochrane risk-of-bias tool; RevMan 5.3, fixed effects when I²<50%, otherwise random effects.
Adults with chronic plantar fasciitis in 9 sham-controlled RCTs (1996–2015), 935 patients: 6 trials of focused and 3 classed as radial shockwaves. Trials giving local anaesthesia to all patients were excluded. Group mean ages 40 to 58.9 years; follow-up 6 weeks to 6 months.
Focused shockwaves (0.06–0.25 mJ/mm², 1,000–2,100 pulses, 3 sessions) and shockwaves classed as radial (0.16 mJ/mm², 2,000 pulses, 2–3 sessions; one trial a single 3,800-pulse session). Two classifications are open to question. Malay 2006 describes an electrohydraulic device and Marks 2008 a pneumatic one.
Sham treatments: air-filled or foil standoffs, foam-insulated membrane, sound-reflecting pad, zero- or minimal-energy pulses, a heel clasp, a placebo handpiece, or the applicator held 1 cm away.
Treatment success as defined by each trial (mostly >50% or >60% VAS improvement, or VAS <4 cm), pooled as odds ratios; reduction in pain scale in 4 trials as standardized mean differences; complications. Follow-up 6 weeks to 6 months; function was not analysed.
Success: all shockwaves OR 2.58 (95% CI 1.97–3.39), 9 trials, I²=38%; focused OR 2.17 (1.49–3.16), 6 trials, I²=0%; radial OR 4.63 (1.30–16.46), 3 trials, I²=81%, and OR 2.39 (1.57–3.65) without the outlying Ibrahim trial; Gerdesmeyer 2008 alone OR 2.11 (1.27–3.52). Pain reduction: SMD 1.01 (−0.01 to 2.03), p=.05, I²=96%; focused 1.29 (0.39–2.19), I²=87%; without Gollwitzer 2015, 0.57 (0.12–1.01). Two focused trials, Speed 2003 and Marks 2008, showed no significant difference on their own.
Seven of 9 trials described complications. No serious adverse events (such as haematoma or infection) were recorded; a few patients had discomfort, pain, swelling or bruising during or after treatment. No pooled harm estimate was made, and two trials did not describe complications, so the authors’ statement that ESWT is safe rests on limited reporting.
The title (“is effective”) goes beyond the conclusion, which reserves judgment on ESWT overall and on radial therapy. The text mislabels treated-arm totals as events; Table 1 and Figure 4 counts disagree for two trials; forest-plot axis labels run opposite to the data; and values described as standard errors may have been used as SDs. Malay 2006 (electrohydraulic) is analysed as radial and Marks 2008 (pneumatic) as focused. English-only search; high heterogeneity for radial and pain outcomes; function not analysed. Government and hospital funding; no conflicts declared.
Applies to adults with chronic plantar fasciitis treated without routine local anaesthesia and compared with sham over at most 6 months. The focused-device success result is the most consistent; the radial and pain-scale figures need verification, and the review cannot say which device type works better.
Reanalyse with verified arm counts, correct device classification and consistent SD/SE handling; trials comparing focused and radial devices directly with function outcomes and longer follow-up.
Read the original article.
Extracorporeal shock wave therapy is effective in treating chronic plantar fasciitis: A meta-analysis of RCTs.
Sun J, Gao F, Wang Y, Sun W, Jiang B, Li Z. · Medicine (Baltimore) · 2017
Open the publication record or available full-text files.
Checked against the full paper · Clinical review: Dr. Marcus Yu Bin Pai, · How we review →