Chronic plantar fasciitis: one electrohydraulic session versus sham · Randomized trial · 2006
Does one shockwave session help long-lasting heel pain?
In this trial, one session with this device lowered heel pain a little more than a sham session. At 3 months, pain fell by about 2.5 points vs about 1.6 with sham, on a 0–10 scale. At 1 month the gap was about 0.3 points and not statistically clear, and walking and activity were not clearly better.
Helped a little more than sham at 3 months- adults with long-lasting heel pain
- 172
- session of about 25 minutes
- 1
- of compared results
- 3 mo
Good to know: Funded by the device maker.
Malay DS, Pressman MM, Assili A, et al. J Foot Ankle Surg, 2006 · Read the paper ↗
How this study worked
What happened
Heel pain under pressure: extra drop vs sham
Difference in pain change, 0–10 scale, with likely range · 0 means no difference
Show the numbers
| Difference (points on the 0–10 pain scale) | likely range | |
|---|---|---|
| 1 month | −0.34 | −1.06 to 0.37 |
| 2 months | −0.99 | −1.86 to −0.12 |
| 3 months | −0.94 | −1.87 to −0.02 |
The treatment
Orthospec active shockwaves
Shockwave, 1 session
- Sessions
- 1
- Pulses
- 3,800 per session · strength raised step by step from level 1 to 7, as tolerated
- Numbing
- None
- Device
- Orthospec (Medispec)
- Everyone also got
- No other treatment; pain medicines paused for 48 hours before each visit
Compared with: A sham session: the same procedure, with foam blocking most of the energy
Settings are what each study used, not a prescription.
Who took part
172 adults with pain in one heel for over 6 months despite several treatments, at three US centres; average age 51.
Side effects
3 of 115 treated people had a mild bruise or swelling that settled within days; none in the sham group.
All the numbers and the full appraisal
The results, in numbers.
172 adults with proximal plantar fasciitis for over 6 months despite at least 4 months of professional care and two drug plus two non-drug treatments; heel pain ≥5 cm on the VAS; one heel per person, bilateral cases excluded. Mean age 51 years, 33% men, 87% white, mean pain duration 30 months; three US centres, 2003–2004.
People, one heel each. 112 entered the 3-month LOCF efficacy analysis; observed visits: 111 at 1 month, 97 at 2, 101 at 3, 87 at 6 and 78 at 12 months. One person randomized to ESWT received sham by mistake; this participant was placed in the placebo group for the energy analyses (Figure 4, Table 10, Figures 5–6).
People, one heel each. 56 entered the 3-month LOCF efficacy analysis; observed visits: 54 at 1 month, 48 at 2, 51 at 3, 45 at 6 and 40 at 12 months.
Investigator-assessed heel pain (primary)5 reported outcomes
Investigator-assessed heel pain (primary)
Investigator VAS with PressureSpec standardized pressure VAS points (cm)
| Group | n | Baseline | Reported change |
|---|---|---|---|
| Orthospec active shockwaves | 112 | — | -2.51 |
| Foam-insulated sham | 56 | — | -1.57 |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Active shockwaves versus sham
Means from analysis of variance adjusted for baseline score and clinical site; last observation carried forward
Protocol-defined primary outcome. Pain fell in both groups; the reduction was 0.94 points larger with active treatment, with a confidence interval reaching −0.02, close to no difference.
Analysis details and source
Adjusted mean change from baseline; negative values mean less pain. SDs of the change were not reported. Analysis included people with at least one follow-up visit, with LOCF.
Investigator-assessed heel pain
Investigator VAS with PressureSpec standardized pressure VAS points (cm)
| Group | n | Baseline | Reported change |
|---|---|---|---|
| Orthospec active shockwaves | 111 | — | -1.61 |
| Foam-insulated sham | 54 | — | -1.27 |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Active shockwaves versus sham
Means from analysis of variance adjusted for baseline score and clinical site; last observation carried forward
At 1 month the difference was small and not statistically significant (−0.34, 95% CI −1.06 to 0.37).
Analysis details and source
Adjusted mean change from baseline; negative values mean less pain. SDs of the change were not reported. Analysis included people with at least one follow-up visit, with LOCF.
Investigator-assessed heel pain
Investigator VAS with PressureSpec standardized pressure VAS points (cm)
| Group | n | Baseline | Reported change |
|---|---|---|---|
| Orthospec active shockwaves | 111 | — | -2.3 |
| Foam-insulated sham | 54 | — | -1.31 |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Active shockwaves versus sham
Means from analysis of variance adjusted for baseline score and clinical site; last observation carried forward
At 2 months the reduction was larger with active treatment.
Analysis details and source
Adjusted mean change from baseline; negative values mean less pain. SDs of the change were not reported. Analysis included people with at least one follow-up visit, with LOCF.
Investigator pain, no heel spur on X-ray
Investigator VAS with PressureSpec standardized pressure VAS points (cm)
| Group | n | Baseline | Reported change |
|---|---|---|---|
| Orthospec active shockwaves | 68 | — | -3.67 |
| Foam-insulated sham | 36 | — | -2.19 |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Active shockwaves versus sham
Means from analysis of variance adjusted for baseline score and clinical site; last observation carried forward
Without a radiographic spur the reduction was 1.48 points larger with active treatment.
Analysis details and source
Subgroup defined by baseline radiograph (68 versus 36 people); no treatment-by-spur interaction test is reported.
Investigator pain, heel spur on X-ray
Investigator VAS with PressureSpec standardized pressure VAS points (cm)
| Group | n | Baseline | Reported change |
|---|---|---|---|
| Orthospec active shockwaves | 45 | — | -2.06 |
| Foam-insulated sham | 20 | — | -1.99 |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Active shockwaves versus sham
Means from analysis of variance adjusted for baseline score and clinical site; last observation carried forward
With a radiographic spur, pain fell in both groups and the difference was small and not statistically significant (−0.07, 95% CI −1.66 to 1.52). The authors say a larger sample would be needed to judge this subgroup.
Analysis details and source
Subgroup of 45 versus 20 people; wide interval. The Summary nonetheless presents the device as an option with or without a spur.
Self-assessed heel pain3 reported outcomes
Self-assessed heel pain
Participant VAS, pain on first steps in the morning VAS points (cm)
| Group | n | Baseline | Reported change |
|---|---|---|---|
| Orthospec active shockwaves | 110 | — | -2.23 |
| Foam-insulated sham | 54 | — | -2.12 |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Active shockwaves versus sham
Means from analysis of variance adjusted for baseline score and clinical site; last observation carried forward
Self-rated pain fell in both groups at 1 month; the difference was not statistically significant (−0.11, 95% CI −0.95 to 0.72).
Analysis details and source
Adjusted mean change from baseline; negative values mean less pain. SDs of the change were not reported. Analysis included people with at least one follow-up visit, with LOCF.
Self-assessed heel pain
Participant VAS, pain on first steps in the morning VAS points (cm)
| Group | n | Baseline | Reported change |
|---|---|---|---|
| Orthospec active shockwaves | 111 | — | -2.67 |
| Foam-insulated sham | 54 | — | -1.94 |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Active shockwaves versus sham
Means from analysis of variance adjusted for baseline score and clinical site; last observation carried forward
At 2 months the difference in self-rated pain was not statistically significant.
Analysis details and source
Adjusted mean change from baseline; negative values mean less pain. SDs of the change were not reported. Analysis included people with at least one follow-up visit, with LOCF.
Self-assessed heel pain
Participant VAS, pain on first steps in the morning VAS points (cm)
| Group | n | Baseline | Reported change |
|---|---|---|---|
| Orthospec active shockwaves | 112 | — | -3.39 |
| Foam-insulated sham | 56 | — | -1.78 |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Active shockwaves versus sham
Means from analysis of variance adjusted for baseline score and clinical site; last observation carried forward
At 3 months self-rated pain fell 1.61 points more with active treatment.
Analysis details and source
Adjusted mean change from baseline; negative values mean less pain. SDs of the change were not reported. Analysis included people with at least one follow-up visit, with LOCF.
Investigator-rated responders3 reported outcomes
Investigator-rated responders
Investigator VAS responder people meeting the responder definition
Active shockwaves versus sham
Logistic regression including baseline value and clinical site; LOCF
Logarithmic ratio scale.
At 1 month the difference was not statistically significant (OR 1.30, 0.52–3.27).
Analysis details and source
Responder: ≥50% decrease from baseline and a score ≤4 cm. Missing visits were imputed with LOCF; the paper reports these as n/N.
Investigator-rated responders
Investigator VAS responder people meeting the responder definition
Active shockwaves versus sham
Logistic regression including baseline value and clinical site; LOCF
Logarithmic ratio scale.
More responders with active treatment at 2 months.
Analysis details and source
Responder: ≥50% decrease from baseline and a score ≤4 cm. Missing visits were imputed with LOCF; the paper reports these as n/N.
Investigator-rated responders
Investigator VAS responder people meeting the responder definition
Active shockwaves versus sham
Logistic regression including baseline value and clinical site; LOCF
Logarithmic ratio scale.
At 3 months 48 of 112 versus 11 of 56 met the responder definition.
Analysis details and source
Responder: ≥50% decrease from baseline and a score ≤4 cm. Missing visits were imputed with LOCF; the paper reports these as n/N.
Self-rated responders3 reported outcomes
Self-rated responders
Participant VAS responder people meeting the responder definition
Active shockwaves versus sham
Logistic regression including baseline value and clinical site; LOCF
Logarithmic ratio scale.
At 1 month the difference was not statistically significant (OR 1.20, 0.60–2.41).
Analysis details and source
Responder: ≥50% decrease from baseline and a score ≤4 cm. Missing visits were imputed with LOCF; the paper reports these as n/N.
Self-rated responders
Participant VAS responder people meeting the responder definition
Active shockwaves versus sham
Logistic regression including baseline value and clinical site; LOCF
Logarithmic ratio scale.
The 2-month difference was not statistically significant.
Analysis details and source
Responder: ≥50% decrease from baseline and a score ≤4 cm. Missing visits were imputed with LOCF; the paper reports these as n/N.
Self-rated responders
Participant VAS responder people meeting the responder definition
Active shockwaves versus sham
Logistic regression including baseline value and clinical site; LOCF
Logarithmic ratio scale.
At 3 months 59 of 112 versus 16 of 56 met the responder definition.
Analysis details and source
Responder: ≥50% decrease from baseline and a score ≤4 cm. Missing visits were imputed with LOCF; the paper reports these as n/N.
Activity and function responders3 reported outcomes
Activity and function responders
Participant diary activity and function people meeting the responder definition
Active shockwaves versus sham
Logistic regression including baseline value and clinical site; LOCF
Logarithmic ratio scale.
At 1 month the difference was not statistically significant (OR 0.82, 0.41–1.66).
Analysis details and source
Responder: improvement of ≥1 point or maintenance of a baseline score of 0 or 1 in the diary activity/function category; LOCF.
Activity and function responders
Participant diary activity and function people meeting the responder definition
Active shockwaves versus sham
Logistic regression including baseline value and clinical site; LOCF
Logarithmic ratio scale.
The 2-month difference was not statistically significant.
Analysis details and source
Responder: improvement of ≥1 point or maintenance of a baseline score of 0 or 1 in the diary activity/function category; LOCF.
Activity and function responders
Participant diary activity and function people meeting the responder definition
Active shockwaves versus sham
Logistic regression including baseline value and clinical site; LOCF
Logarithmic ratio scale.
At 3 months 72 of 112 versus 32 of 56 met the activity and function definition; the difference was not statistically significant.
Analysis details and source
Responder: improvement of ≥1 point or maintenance of a baseline score of 0 or 1 in the diary activity/function category; LOCF.
Decreased analgesic use1 reported outcome
Decreased analgesic use
Participant diary, change in pain medication since enrolment people reporting decreased use
Active versus sham across increase, no change and decrease
Cochran–Mantel–Haenszel, stratified by site, scores 1/0/−1
At 3 months 34/100 active versus 7/51 sham reported less analgesic use and 1/100 versus 6/51 reported more. Month 2: decrease 28/95 versus 9/47 (p=.028); month 1: p=.083.
Analysis details and source
Observed diary denominators (100 and 51), not imputed. The test compares all three categories (increase, no change, decrease), not the decrease proportion alone.
Energy level above 4.5 (post hoc)1 reported outcome
Energy level above 4.5 (post hoc)
qualitative finding
Restricting the active group to people who tolerated level >4.5 (Table 10: 88 versus 52 placebo), investigator pain change was −2.98 versus −1.69 (p=.011) and self-rated −3.95 versus −1.82 (p<.001); activity and function response 67.0% versus 56.1% (p=.16).
Analysis details and source
Post hoc subset chosen by tolerated energy after randomization, so it is not a randomized comparison. The 88 and 52 denominators are not explained, and Figure 5 reports a placebo month-3 change of 1.96, not 1.69.
Sex, age and weight subgroups1 reported outcome
Sex, age and weight subgroups
qualitative finding
Interaction tests for investigator pain at 3 months were not significant for sex (p=.15), age (p=.89) or weight (p=.35). Differences ranged from −0.21 (161–192 lb) to −1.91 (men; ≥193 lb).
Analysis details and source
Nonsignificant interaction tests in small subgroups do not show that the effect is the same in each subgroup, as the Discussion implies.
How treatment was delivered
Orthospec (Medispec): electrohydraulic spark-gap generator in a water-filled semi-ellipsoid chamber, delivered through a rubber contact membrane to a 35-mm therapy zone. The paper does not call it focused or radial and gives no energy flux density in mJ/mm² for its levels.
One 25-minute session; 3,800 shocks at 150 per minute, delivered continuously from energy level 1, rising every 3 min 35 s to level 7; the unblinded operator lowered the level if not tolerated. Highest level reached among 115: 2–4.5 in 14, 4.6–5.9 in 12, 6–6.9 in 54, ≥7 in 35.
No local or systemic anaesthesia or sedation; ultrasound gel on membrane and heel; no ultrasonic or radiographic targeting.
Foam-insulated contact membrane to absorb the shockwaves and block most of the energy; same appearance and identical procedure, with the audible sound; ear protection offered. The authors state the sham was not a pure placebo.
Before treatment: 6 weeks without corticosteroid injection, 48 hours without NSAIDs or analgesics, 2 weeks without physical therapy. People continuing physical therapy or other conservative care were excluded; NSAIDs or analgesics were avoided for 48 hours before each follow-up visit.
Blinded investigator assessment: standardized pressure on the heel with the calibrated PressureSpec applicator, pain marked on a 10-point VAS; participants also rated first-steps pain; activity (blocks walked) and analgesic use from diaries. Efficacy visits at 1, 2 and 3 months; safety visits at 6 and 12 months.
No serious adverse events. No placebo participant (N=57) had an adverse event. Three of 115 active participants had one event each: two bruising at the treated heel (1.7%), judged device-related, and one local swelling (0.9%), judged unrelated; energy levels 4.5–7. The Results first describe all three events as at least possibly device-related, then classify the swelling as unrelated. The Results call them mild or moderate, the Discussion mild and resolved within days. Safety visits ran to 12 months, but only 78 and 40 people attended then.
Funding and role of sponsor: Medispec, the device maker, sponsored this FDA investigational-device phase-3 trial, provided the shockwave and pressure devices, funded the investigation, and a Medispec employee did site management and monitoring. Author conflicts are not otherwise stated.
Primary result is small and borderline (−0.94, 95% CI −1.87 to −0.02); at 1 month the difference was not statistically significant (−0.34, −1.06 to 0.37), and activity/function differences were not statistically significant at any month (3 months OR 1.42, 0.70–2.85). The Summary statement that all heel-pain assessments were significant does not hold for month 1.
Missing data: 172 randomized, 168 analysed with at least one follow-up visit, 152 completed the 12-week efficacy period; LOCF was used. Randomization planned 183 people (122 versus 61) for 80% power; 172 were randomized.
Internal inconsistencies: Results say 89 of 115 tolerated energy level ≥6, the Discussion says the same 89 tolerated ≥4.5; Figure 5 placebo month-3 change 1.96 versus Table 10 −1.69; Table 10 denominators (88, 52) unexplained. Blinding was assessed after the session but the result is not reported.
Author-reported limits: no quality-of-life or foot-specific instrument; efficacy measured only to 3 months; diaries open to bias; the sham may have delivered some energy; bilateral cases excluded, limiting generalizability.
Extractor note: the Orthospec is described as an electrohydraulic semi-ellipsoid reflector device; the paper does not call it radial. Some later reviews (Sun 2017) pool this trial as radial shockwave therapy.
Findings checked against the full source by a separate automated reviewer.
Limitations noted by the reviewers
- Funding and role of sponsor: Medispec, the device maker, sponsored this FDA investigational-device phase-3 trial, provided the shockwave and pressure devices, funded the investigation, and a Medispec employee did site management and monitoring. Author conflicts are not otherwise stated.
- Primary result is small and borderline (−0.94, 95% CI −1.87 to −0.02); at 1 month the difference was not statistically significant (−0.34, −1.06 to 0.37), and activity/function differences were not statistically significant at any month (3 months OR 1.42, 0.70–2.85). The Summary statement that all heel-pain assessments were significant does not hold for month 1.
- Missing data: 172 randomized, 168 analysed with at least one follow-up visit, 152 completed the 12-week efficacy period; LOCF was used. Randomization planned 183 people (122 versus 61) for 80% power; 172 were randomized.
- Internal inconsistencies: Results say 89 of 115 tolerated energy level ≥6, the Discussion says the same 89 tolerated ≥4.5; Figure 5 placebo month-3 change 1.96 versus Table 10 −1.69; Table 10 denominators (88, 52) unexplained. Blinding was assessed after the session but the result is not reported.
- Author-reported limits: no quality-of-life or foot-specific instrument; efficacy measured only to 3 months; diaries open to bias; the sham may have delivered some energy; bilateral cases excluded, limiting generalizability.
- Extractor note: the Orthospec is described as an electrohydraulic semi-ellipsoid reflector device; the paper does not call it radial. Some later reviews (Sun 2017) pool this trial as radial shockwave therapy.
Source-linked critical reading
Randomized trial
Technical source record
Complete local PDF read (all 15 pages): methods, results, all 13 tables and 7 figures, discussion, disclosure, acknowledgments and references. No supplement or protocol document was available.
In adults with long-standing heel pain in one foot, a single session with this device lowered pain slightly more than a sham session at 3 months, mainly by self-rating; at 1 month the difference was not statistically significant and walking activity was not clearly better. The trial was run for device approval and funded by the manufacturer.
Does one session with the Orthospec electrohydraulic shockwave device reduce chronic proximal plantar heel pain more than a sham session over 3 months?
Randomized 2:1, double-blind, sham-controlled trial at three US centres (October 2003 to December 2004), run as an FDA investigational-device phase-3 study sponsored by the device maker. 172 randomized (115 active, 57 sham); 168 with at least one follow-up analysed with last observation carried forward.
172 adults with proximal plantar fasciitis for over 6 months despite at least 4 months of professional care and two drug plus two non-drug treatments; heel pain ≥5 cm on the VAS; one heel per person, bilateral cases excluded. Mean age 51 years, 33% men, 87% white, mean pain duration 30 months; three US centres, 2003–2004.
Single 25-minute session: 3,800 shocks at 150 per minute from an electrohydraulic spark-gap device with a 35-mm treatment zone, energy raised from level 1 to 7 as tolerated, no anaesthesia and no imaging guidance. Energy levels are not given in mJ/mm². 89 of 115 reached level 6 or higher.
The same procedure with a foam-insulated membrane that absorbed most of the energy and looked identical. Recent steroid injections, NSAIDs and physical therapy were stopped beforehand, and other heel treatments were not continued.
Primary: change in blinded investigator-rated heel pain (standardized pressure, 10-point VAS) from baseline to 3 months. Secondary: self-rated first-steps pain, responder rates (≥50% fall and ≤4 cm), diary activity and function, and analgesic use at 1, 2 and 3 months; adverse events to 12 months.
At 3 months investigator-rated pain fell 2.51 versus 1.57 points, difference −0.94 (95% CI −1.87 to −0.02), p=.045; at 1 month −0.34 (−1.06 to 0.37). Self-rated pain at 3 months: −3.39 versus −1.78, difference −1.61 (−2.55 to −0.67); at 1 and 2 months the differences were not significant. Responders at 3 months: 48/112 versus 11/56 by investigator and 59/112 versus 16/56 by self-rating. Activity and function differences were not statistically significant at any month (3 months 72/112 versus 32/56, OR 1.42, 0.70–2.85, p=.33). With a heel spur the pain difference was −0.07 (−1.66 to 1.52).
No serious adverse events. No placebo participant (N=57) had an adverse event. Three of 115 active participants had one event each: two bruising at the treated heel (1.7%), judged device-related, and one local swelling (0.9%), judged unrelated; energy levels 4.5–7. The Results first describe all three events as at least possibly device-related, then classify the swelling as unrelated. The Results call them mild or moderate, the Discussion mild and resolved within days. Safety visits ran to 12 months, but only 78 and 40 people attended then.
The device maker sponsored and funded the trial, supplied devices and monitored sites. The primary difference was small, its interval nearly reached zero, and LOCF filled missing visits. Energy-level analyses were post hoc and nonrandomized, with unexplained denominators and a Figure 5 versus Table 10 mismatch (1.96 versus −1.69). The Results and Discussion disagree on the energy level the same 89 people tolerated (≥6 versus ≥4.5). Blinding was assessed but not reported, and the authors note the sham may have delivered some energy. No quality-of-life measure; efficacy only to 3 months; bilateral cases excluded.
Applies to adults with one painful heel for over 6 months after several conservative treatments, treated with this specific electrohydraulic device without anaesthesia. It suggests a modest pain difference at 3 months; it does not show better walking function, a benefit in people with heel spurs, or what happens after 3 months. Dose findings by energy level are hypothesis-generating.
Independent trials with a checked sham, analysis of every randomized participant, function and quality-of-life outcomes, and efficacy follow-up beyond 3 months.
Read the original article.
Extracorporeal shockwave therapy versus placebo for the treatment of chronic proximal plantar fasciitis: results of a randomized, placebo-controlled, double-blinded, multicenter intervention trial.
Malay DS, Pressman MM, Assili A, Kline JT, York S, Buren B, Heyman ER, Borowsky P, LeMay C. · J Foot Ankle Surg · 2006
Open the publication record or available full-text files.
Checked against the full paper · Clinical review: Dr. Marcus Yu Bin Pai, · How we review →