Plantar heel pain / plantar fasciopathy · Randomized trial · 2004
Does a single focused shockwave session ease morning heel pain?
In this trial, morning heel pain fell by about 4.4 points out of 10 three months after one focused shockwave session, and by about 3.6 after a sham session. Reaching at least 60% less pain was not clearly more common (56% vs 45%). The treatment was usually painful despite a numbing injection.
Helped a little more than sham at 3 months- adults with long-lasting heel pain
- 150
- session, with a numbing injection
- 1
- of compared results
- 3 mo
Good to know: Funded by the device maker.
Theodore GH, Buch M, Amendola A, et al. Foot & Ankle International, 2004 · Read the paper ↗
How this study worked
What happened
Change in morning heel pain
Change on a 0–10 pain scale over three months
Show the numbers
| Group | People | Change (points on the 0–10 pain scale) |
|---|---|---|
| Focused shockwave | 73 | −4.4 |
| Sham | 73 | −3.6 |
The treatment
Focused shockwave
Focused, 1 session
- Sessions
- 1
- Pulses
- 3,800 per session
- Strength
- 0.36mJ/mm² for 3,500 of the shocks
- Numbing
- Local anaesthetic
- Device
- Dornier Epos Ultra
- Everyone also got
- No other treatments during the study
Compared with: A sham session, with an air cushion blocking the shockwaves
Settings are what each study used, not a prescription.
Who took part
150 adults with pain in one heel for more than 6 months after several treatments, at six sites; mostly women, average age about 50.
hollow squares: no final score
Side effects
Most felt pain during treatment (73% vs 7% with sham). Swelling, bruising and tingling were similar in both groups and usually went within a week; in one person tingling was recorded as permanent.
All the numbers and the full appraisal
The results, in numbers.
150 adults (109 women, 41 men) at six clinical sites with unilateral plantar medial heel pain for more than 6 months, morning first-steps pain >5 on a 0–10 cm VAS and Roles and Maudsley 3–4, after at least 6 months of failed NSAIDs plus two other therapies. Mean age 50 versus 53 years; mean symptom duration 22 versus 24 months. Diabetes, neuropathy, bilateral symptoms and recent steroid injection were excluded.
People with unilateral symptoms. 3 discontinued before 3 months; VAS n 74 at 3–5 days, 72 at 6 weeks, 73 at 3 months, then 58 at 6 months and 50 at 12 months (uncontrolled). Repeated-measures analysis "on completed patients"; intention-to-treat not stated.
People with unilateral symptoms. 1 discontinued before 3 months; VAS n 74, 71 and 73 at 3–5 days, 6 weeks and 3 months. Unblinded at 3 months and offered active treatment, so no controlled data afterwards. Some tables use 73 as the baseline denominator.
Morning first-steps pain change (primary)2 reported outcomes
Morning first-steps pain change (primary)
VAS, pain during the first few minutes of walking in the morning VAS points, change from baseline
| Group | n | Baseline | Reported change |
|---|---|---|---|
| Focused ESWT (Epos Ultra) | 73 | — | -4.4 (SD 2.8) |
| Sham (air-cushion blocked) | 73 | — | -3.6 (SD 3.1) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Active ESWT versus sham
ANCOVA on change from baseline (fixed effects for treatment and site; covariates baseline VAS, body weight, symptom duration)
Pain fell in both groups, from 7.7 to 3.4 with active treatment and from 7.7 to 4.1 with sham; the adjusted between-group test favoured active treatment (P=.0435). A repeated-measures analysis through 3 months on completed patients gave P=.0149. Mean percentage improvements were reported as 57% versus 47%.
Analysis details and source
No effect estimate or CI for the difference. Table 4 gives 3-month SD 2.7 for the active group while the text writes 3.4 ± 2.8. The P=.0001 values in Table 4 are within-group paired t tests.
Morning first-steps pain at 6 weeks
VAS, pain during the first few minutes of walking in the morning VAS points (0–10 cm scale)
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Focused ESWT (Epos Ultra) | 72 | 7.7 (SD 1.4) | 4.6 (SD 3.1) |
| Sham (air-cushion blocked) | 71 | 7.7 (SD 1.5) | 5 (SD 3) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Mean pain was 4.6 versus 5.0 at 6 weeks (5.0 versus 5.7 at 3–5 days). No between-group test is reported for these visits.
Analysis details and source
Descriptive values; baseline n 76 and 74.
Clinical success: ≥60% less morning pain1 reported outcome
Clinical success: ≥60% less morning pain
At least 60% improvement in first-steps VAS People
Active ESWT versus sham
Between-group comparison of proportions (test not named)
56% (41/73) versus 45% (33/73); the difference was not statistically significant (P=.1885).
Analysis details and source
The abstract reports "56% success" for active and "47% success" for control without saying the success comparison was nonsignificant; 47% is the control group’s mean percentage VAS improvement, and its success rate was 45%.
Roles and Maudsley excellent or good1 reported outcome
Roles and Maudsley excellent or good
Roles and Maudsley Score (four-point patient self-assessment) People
Active ESWT versus sham
Between-group test as printed in Table 6 (method not named)
45/73 (61.6%) versus 29/73 (39.7%) rated excellent or good at 3 months (P=.0327); at 6 weeks 25/72 versus 23/71 (P=.9343).
Analysis details and source
Table 6 shows 1/76 and 1/73 excellent-good at baseline although inclusion required a score of 3 or 4, prints the control fair-poor share at 3 months as 44/73 (6.3%), and prints 25/72 as 35.7%.
Pain change with symptoms over 12 months2 reported outcomes
Pain change with symptoms over 12 months
VAS, first steps in the morning VAS points, change from baseline
| Group | n | Baseline | Reported change |
|---|---|---|---|
| Focused ESWT (Epos Ultra) | — | — | -5.1 |
| Sham (air-cushion blocked) | — | — | -3.7 |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Active ESWT versus sham
Between-group comparison within the subgroup (test not named)
Among people with symptoms for more than 12 months, average change was −5.1 versus −3.7 (P=.0309).
Analysis details and source
Subgroup sizes, SDs and whether the subgroup was prespecified are not reported; treat as exploratory.
Morning pain at 12 months, active group only
VAS, first steps in the morning VAS points (0–10 cm scale)
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| Focused ESWT (Epos Ultra) | 50 | 7.7 (SD 1.4) | 0.6 (SD 1.2) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
In the 50 active-group patients seen at 12 months, mean pain was 0.6 (91.3% change from baseline); at 6 months it was 2.2 in 58. With no control group after 3 months, this does not show what shockwaves added.
Analysis details and source
Uncontrolled follow-up; 26 of 76 not represented at 12 months. The abstract’s "94% success" at 12 months is from the active group only, with no control group, and its definition is not shown; Roles and Maudsley excellent-good was 48/51 (94.1%) and the VAS change 91.3% in 50 patients.
AOFAS and SF-12 scores1 reported outcome
AOFAS and SF-12 scores
AOFAS Ankle-Hindfoot Scoring System; SF-12 Not reported
No statistically significant between-group differences; follow-up values were not reported, only numerical trends favouring active treatment in AOFAS pain and SF-12 physical scores.
Analysis details and source
Baseline values are in Table 3 only.
Pain during treatment1 reported outcome
Pain during treatment
Adverse event: pain during shock wave application People
Active ESWT versus sham
Fisher exact test
55 of 76 (73%) felt pain during active treatment versus 5 of 74 (7%) with sham, despite the nerve block.
Analysis details and source
Table 7 prints P<.001 and its footnote P<.0001. The difference may have revealed allocation; blinding success was not reported.
Crossover group after active treatment1 reported outcome
Crossover group after active treatment
VAS and Roles and Maudsley in sham patients later treated Percent change; people
Sham patients treated after unblinding had a 53% VAS change at 3 months (n=38) and 76% at 12 months (n=30); 28/36 and 27/29 rated good to excellent. This is uncontrolled, open follow-up.
Analysis details and source
Baseline was the original 3-month value; no comparison group.
How treatment was delivered
Dornier Epos Ultra electromagnetic focused shockwave system with in-line (isocentric) ultrasound; source position adjusted during treatment by ultrasound image and patient feedback to the pain epicentre. Single session, patients prone.
3,800 shocks, 3,500 of them at 0.36 mJ/mm², for a stated total of 1,300 mJ/mm². Shock frequency and the energy-flux definition (ED or ED+) are not reported.
Medial calcaneal nerve block with 5 mL of 1% xylocaine 15–20 minutes before the procedure, in both groups.
Identical process with a thin air cushion on the therapy head to prevent shock wave penetration; the cushion was fitted before the patient entered the room.
Participants agreed to forgo other therapies during the study. Blinded physician assessments at 3–5 days, 6 weeks and 3 months; unblinding at 3 months, with active treatment offered to unimproved sham patients (crossover group); active patients followed to 12 months.
Through 3 months (Table 7, 76 versus 74): pain during treatment 55 (73%) versus 5 (7%); pain after treatment 28 (37%) versus 24 (32%); edema 5 versus 6; ecchymosis 5 versus 4; hypesthesia 2 versus 6; paresthesia 3 versus 3; rash 1 versus 0; neuralgia 1 versus 0; petechiae 0 versus 1. Common events resolved within a week; one patient withdrew before paresthesia resolved and it was coded permanent. Total events 104 versus 53. All sites received grants from Dornier MedTech America, the device maker and study sponsor.
Author-reported: placebo improvement may reflect placebo response or natural history; the study was not sized for AOFAS and SF-12 differences.
Extractor interpretation: the abstract presents 56% versus 47% "success" and 94% at 12 months without stating that the prespecified success comparison was nonsignificant (P=.1885) or that 12-month data were uncontrolled.
Extractor interpretation: pain during treatment (73% versus 7%) may unmask allocation for a self-reported outcome, and blinding was not checked; the primary analysis used completed patients, with no estimate or CI for the difference.
Reporting: groups differed in sex (female P=.0156) and height (P=.0131); several printed ranges in Tables 2–3 are implausible (height 6.4–77.0 inches, weights 12.0 and 39.0 lb, VAS 5.0–1.0, control symptom duration from 3.0 months against a >6-month inclusion rule); Table 6 has percentage misprints. The device maker sponsored the trial.
Findings checked against the full source by a separate automated reviewer.
Limitations noted by the reviewers
- Author-reported: placebo improvement may reflect placebo response or natural history; the study was not sized for AOFAS and SF-12 differences.
- Extractor interpretation: the abstract presents 56% versus 47% "success" and 94% at 12 months without stating that the prespecified success comparison was nonsignificant (P=.1885) or that 12-month data were uncontrolled.
- Extractor interpretation: pain during treatment (73% versus 7%) may unmask allocation for a self-reported outcome, and blinding was not checked; the primary analysis used completed patients, with no estimate or CI for the difference.
- Reporting: groups differed in sex (female P=.0156) and height (P=.0131); several printed ranges in Tables 2–3 are implausible (height 6.4–77.0 inches, weights 12.0 and 39.0 lb, VAS 5.0–1.0, control symptom duration from 3.0 months against a >6-month inclusion rule); Table 6 has percentage misprints. The device maker sponsored the trial.
Source-linked critical reading
Sham-controlled double-blind randomized trial
Technical source record
Full local PDF read: all 8 pages including Tables 1–7 from page images, Methods, Results, Discussion, Acknowledgments (FDA approval and funding) and references.
In this sponsor-funded multicentre trial, adults with long-standing heel pain on one side had a somewhat larger fall in morning pain three months after one high-energy focused shockwave session than after a sham session. Sham patients also improved, the share reaching at least 60% less pain did not differ significantly (56% versus 45%), and treatment was usually painful. The high 12-month figures come from the treated group alone, without a comparison group.
For chronic plantar fasciitis after failed conservative care, does one high-energy electromagnetic focused shockwave session with the Dornier Epos Ultra reduce morning first-steps pain at 3 months more than a sham session?
Multicentre, randomized 1:1, sham-controlled, double-blind trial at six sites, sponsored by the device maker, which obtained FDA approval before the study began. 150 randomized (76 active, 74 sham); 73 per group assessed at 3 months. Blinded physicians assessed outcomes; unblinding at 3 months, then crossover for unimproved sham patients. Allocation method and concealment are not described.
150 adults (109 women, 41 men) at six clinical sites with unilateral plantar medial heel pain for more than 6 months, morning first-steps pain >5 on a 0–10 cm VAS and Roles and Maudsley 3–4, after at least 6 months of failed NSAIDs plus two other therapies. Mean age 50 versus 53 years; mean symptom duration 22 versus 24 months. Diabetes, neuropathy, bilateral symptoms and recent steroid injection were excluded.
One session with the Dornier Epos Ultra after a medial calcaneal nerve block, patient prone, focus guided by in-line ultrasound and patient feedback: 3,800 shocks, 3,500 at 0.36 mJ/mm², stated total 1,300 mJ/mm². Frequency not reported.
Sham: identical process and nerve block, with a thin air cushion on the therapy head preventing shock wave penetration. Other therapies were to be avoided during the study.
Primary: change in VAS pain during the first minutes of walking in the morning at 3 months. The share with at least 60% improvement (clinical success, used in the sample-size assumptions) was also compared. Secondary: other VAS items, Roles and Maudsley, AOFAS, SF-12 and pressure-threshold tenderness. Visits at 3–5 days, 6 weeks and 3 months; active patients also at 6 and 12 months.
Morning pain changed by −4.4 with active treatment and −3.6 with sham (adjusted P=.0435; repeated measures P=.0149). The share with at least 60% improvement was 41/73 versus 33/73 (P=.1885). Roles and Maudsley excellent-good was 45/73 versus 29/73 (P=.0327). AOFAS and SF-12 did not differ significantly. After 3 months only the treated group was followed.
Through 3 months (Table 7, 76 versus 74): pain during treatment 55 (73%) versus 5 (7%); pain after treatment 28 (37%) versus 24 (32%); edema 5 versus 6; ecchymosis 5 versus 4; hypesthesia 2 versus 6; paresthesia 3 versus 3; rash 1 versus 0; neuralgia 1 versus 0; petechiae 0 versus 1. Common events resolved within a week; one patient withdrew before paresthesia resolved and it was coded permanent. Total events 104 versus 53. All sites received grants from Dornier MedTech America, the device maker and study sponsor.
The abstract’s 56% versus 47% "success" mixes a success rate with a mean improvement and omits the nonsignificant success comparison; its 94% figure comes from uncontrolled 12-month data. Pain during treatment (73% versus 7%) may unmask allocation, blinding was not checked, analysis used completed patients without CIs, groups differed in sex and height, and several table ranges are misprinted. The device maker sponsored the study.
Applies to adults with one chronically painful heel after several conservative treatments, treated once with this device and energy under nerve block. It does not test radial shockwaves, repeated low-energy courses or controlled outcomes beyond 3 months.
An independently funded trial that reports allocation concealment, checks blinding with a sensation-matched sham, analyses everyone randomized with effect estimates and CIs, and keeps a controlled comparison to 12 months.
Read the original article.
Extracorporeal shock wave therapy for the treatment of plantar fasciitis.
Theodore GH, Buch M, Amendola A, Bachmann C, Fleming LL, Zingas C. · Foot & ankle international · 2004
Open the publication record or available full-text files.
Checked against the full paper · Clinical review: Dr. Marcus Yu Bin Pai, · How we review →