982613

research-article2021
FAIXXX10.1177/1071100720982613Foot & Ankle InternationalAbdelkader et al

Article

Foot & Ankle International®

Short- and Intermediate-Term Results of

1­–10
© The Author(s) 2021
Article reuse guidelines:
Extracorporeal Shockwave Therapy for sagepub.com/journals-permissions
DOI: 10.1177/1071100720982613
https://doi.org/10.1177/1071100720982613

Noninsertional Achilles Tendinopathy journals.sagepub.com/home/fai

Nasr Awad Abdelkader, PhD1 , Mohamed Nasser Kise Helmy, BSc PT2,
Nadia Abdelazem Fayaz, PhD1, and Emad S. B. Saweeres,
FRCS (Tr & Orth), MD3

Abstract
Background: Earlier randomized controlled trials (RCTs) reported only midterm (3-4 months) results of extracorporeal
shockwave therapy (ESWT) as a treatment for noninsertional Achilles tendinopathy (NAT). This study compared the
effectiveness of an eccentric loading program followed by stretching exercises combined with ESWT (study group) or
sham ESWT (control group) for treating chronic NAT in both the short and long term.
Methods: This double-blind RCT was conducted between 2018 and 2020. Adult patients with unilateral NAT who failed
standard conservative treatment were randomly allocated to either group. Function and pain were assessed at baseline, 1
month, and 16 months using the Victorian Institute of Sport Assessment–Achilles questionnaire (VISA-A) and visual analog
scale (VAS), respectively. Mixed-design analysis of variance and nonparametric statistics were performed. Twenty-two men
and 28 women aged 18 to 40 years were allocated into 2 equally matched groups.
Results: Function and pain scores in the study group were not significantly different from control group scores at baseline
(VISA-A: 22.2 ± 6.5 vs 21.0 ± 5.2 and VAS: 8 ± 1 vs 8 ± 1, respectively). Both groups significantly improved posttreatment
(VISA-A: 85 ± 6.2 vs 53.4 ± 7.7 and VAS: 1 ± 2 vs 7 ± 2, respectively). At the 16-month follow-up, outcome scores
declined slightly but significantly in the study group (VISA-A: 80 ± 5.3; VAS: 3 ± 2) and improved in the control group
(VISA-A: 67 ± 5.6; VAS: 5 ± 1). However, both groups were significantly better than baseline. At both time points, the
study group had significantly superior scores (statistically and clinically) than the control group (P = .0001).
Conclusions: Combining calf eccentric loading with stretching exercises resulted in significant improvements in the pain
and functional scores in patients with NAT. Adding ESWT to this combined protocol resulted in significantly greater
improvements in both the short and long term.
Level of Evidence: Level I, randomized controlled trial.

Keywords: noninsertional Achilles tendinopathy, extracorporeal shockwave, eccentric loading, stretching exercise, long-
term follow-up

Chronic Achilles tendinopathy (AT) is a common, disabling hamstring muscles.10,23 Presently, heavy-load eccentric
overuse condition of the foot and ankle, especially during calf muscle strengthening exercises, popularized by
walking and running. AT has 2 types: the insertional type Alfredson et al,3 are more widely accepted as the initial
affecting the point of insertion of the Achilles tendon at the intervention for patients with noninsertional AT,7,10,26,27,32
calcaneus and the noninsertional type affecting the area 2 to based on randomized controlled trials (RCTs),23,28,39
6 cm proximal to Achilles insertion.7,10 Several good systematic reviews, and meta-analyses.8,43 However,
reviews exist summarizing the pathophysiology, clinical
and imaging features, and treatment options of AT.7,10,26,27,32 1
Faculty of Physical Therapy, Cairo University, Giza, Egypt
Nonoperative treatments should be attempted for at least 6 2
October 6th Hospital, Ministry of Health, Giza, Egypt
months before surgery is offered. A multimodal regimen 3
Orthopaedic Surgery Department, El-Sahel Teaching Hospital, Cairo,
includes painful activity avoidance, nonsteroidal anti- Egypt
inflammatory drugs (NSAIDs), shoe modifications, ortho-
Corresponding Author:
ses, cryotherapy, nitric oxide, and injections.7,10,26,27,32,43 Emad S. B. Saweeres, FRCS (Tr & Orth), MD, Orthopaedic Surgery
Conventional physical therapy treatment for AT con- Department, El-Sahel Teaching Hospital, Cairo 11697, Egypt.
sists of stretching of the gastrocnemius, soleus, and Email: [email protected]
2 Foot & Ankle International 00(0)

evidence for the superior effectiveness of any particular All patients were diagnosed and referred by a physician
exercise program is limited.8,41,47 according to the following criteria: (1) unilateral noninser-
Low-energy extracorporeal shockwave therapy (ESWT) tional AT, (2) pain at the Achilles tendon for at least 6
is a safe and well-tolerated modality, which could be used as months (chronic AT), and (3) failure of conservative treat-
an alternative to or in combination with exercises. Moreover, ment (eg, NSAIDs, ice, shoe modification, and avoidance
satisfactory evidence (RCTs,9,35,38,39,45 systematic reviews, of overuse activities) for at least 3 months.
and meta-analyses1,13,17,42,43) exists to support its effective- Patients who (1) underwent physical therapy 4 weeks
ness for treating chronic noninsertional AT. Costa et al9 have before enrollment in the study; (2) had taken NSAIDs dur-
reported no alleviation of pain during rest or walking (only ing the previous week; (3) had peritendinous injection of
pain during sports was alleviated) when ESWT was com- corticosteroids or anesthetics within the last 4 weeks pre-
pared to placebo. Although Rompe et al39 have demonstrated ceding the study; (4) had bilateral AT; (5) had ankle osteo-
that both ESWT and eccentric exercises were significantly arthritis, radiculopathy, or systemic neurological diseases
better than the wait-and-see policy, they found that the out- that may cause posterior ankle pain; or (6) had previous
comes of shockwave therapy and exercises were compara- injury or surgical treatments of the ankle were excluded
ble. In their more recent RCT, Rompe et al38 compared from the study.
eccentric loading alone with a combination of eccentric Patients enrolled in the study were randomized into 2
exercises and ESWT and found this combination to be more groups (the study and control groups) using computer-gen-
effective than the exercise program alone in alleviating pain erated numbers in sealed opaque envelopes. Patients in the
and improving function. While the rate of complete recovery study group received ESWT in addition to conservative
using either eccentric loading or ESWT alone was 60% and physical therapy treatments consisting of eccentric training
52%, respectively,39 the combination of the 2 modalities of the calf muscles followed by stretching of the gastrocne-
increased the complete recovery rate to 82%.38 mius, soleus, and hamstring muscles. Patients in the control
Two other RCTs have incorporated stretching exercises group received the same conservative physical therapy
along with eccentric loading in their conservative (control) treatment as well as sham ESWT. The total duration of
program.35,45 Both studies randomized their patients to treatment was 4 weeks. All patients were instructed to avoid
receive either active or sham ESWT combined with the taking NSAIDs or any analgesics during the treatment
aforementioned exercise program. While both trials showed period, and they were blinded to the type of treatment they
a better improvement in the active ESWT group, only received.
improvement in function was statistically significant in the
first study.35 The pain reduction in the first study and the
Baseline
reduction of pain and improvement in function in the other
study45 failed to reach statistical significance. In addition, Seventy-four patients were assessed for eligibility. Twenty-
both studies reported only midterm follow-up outcomes two patients did not meet the inclusion criteria (5 patients
(3-4 months). had bilateral AT, 7 had a history of peritendinous corticoste-
The aim of the present double-blind prospective RCT roid injections, and 10 had an associated ankle pain) and 2
was to compare the effectiveness of a program composed of refused to participate; all were, therefore, excluded (Figure 1).
eccentric training followed by stretching exercises (control The study population consisted of 22 men and 28 women
group) and the same program combined with ESWT (study aged 18 to 40 years with chronic noninsertional AT who
group) for treating chronic noninsertional AT in both the were randomly assigned to 2 equal groups using computer-
short and long term (16-month follow-up). generated numbers in sealed envelopes. The 2 groups were
comparable (Table 1) with no significant differences in any
of the demographic characteristics.
Methods
Study Design Treatment
This double-blind RCT was conducted at the faculty of Extracorporeal shockwave therapy. All patients in the study
physical therapy in Cairo, Egypt, from October 2018 to group received 4 sessions of shockwave at weekly intervals
May 2020. The study sample included 50 patients clini- using an ESWT machine (DOULITH SD1; Storz Medical).
cally diagnosed with noninsertional AT who were referred Each session of ESWT consisted of 2000 pulses with 3 bar
by the orthopedic department. Patients who agreed to pressure (equals an energy flux density of 0.1 mJ/mm²) and
participate signed an informed consent form. The study frequency of 8 pulses/s. The treatment was performed on
was approved by the institutional ethical committee the patients while they were in the prone position with a
(number: P.T.REC/012/002042) of our faculty and regis- small pillow under the ankle (neutral position). Shockwaves
tered at ClinicalTrials.gov (ID: NCT04376294). were applied in a circumferential pattern, targeting the point
Abdelkader et al 3

Figure 1. Flowchart of the trial from the baseline. All patients were assessed posttreatment (at 4 weeks) and at long-term follow-up
(at a minimum period of 13 months).

Table 1. Demographic Characteristics of the Study Cohort.

Items Study group, mean (SD) Control group, mean (SD)

Age, y 29.9 (7.0) 28.3 (6.8)
Body mass, kg 78.2 (10.6) 77.8 (9.2)
Height, cm 172.6 (10.6) 174.4 (8.5)
Body mass index, kg/m2 25.7 (1.6) 25.3 (1.6)

of maximum tenderness.35,38,39 Patients in the control group study, patients in both groups were blinded to the type treat-
were placed in the same position and received sham ESWT. ment they received.
Ultrasound gel was applied to the patient’s skin, and the
handpiece of the machine was moved in a similar fashion. Eccentric training of the calf muscles. Patients were asked to
The machine settings were adjusted to generate zero energy stand on the edge of a wooden step with all their body
while producing the same sound effect. Throughout the weight on the forefoot of the affected leg and the ankle in
4 Foot & Ankle International 00(0)

plantarflexion. The patients were then asked to slowly dor- Each patient was assessed 1 day before starting the first
siflex the ankle to a count of 5, loading the calf muscle session (baseline assessment), 1 day after the last session
eccentrically and lowering the affected limb until the planter (after 4 weeks from baseline; posttreatment), and after a
surface of the heel became lower than the level of the minimum of 13 months from baseline (mean ± SD = 16 ±
wooden step. This maneuver was performed with the knee 2.3 months; range, 13-18 months; follow-up) by an inde-
of the affected lower limb extended to load the gastrocne- pendent assessor (a physiotherapist) who did not administer
mius and with the knee flexed to load the soleus muscle. the ESWT (real or sham) and was only gathering the out-
The calf muscles were loaded only eccentrically, as the come data. This assessor was not involved in the study and
patients were instructed to return to the starting position was blinded to the patients’ allocations to either group.
using the sound limb or his upper limbs. The patients in
both groups were instructed to perform 3 sets of 15 repeti- Statistical analysis. Sample size was calculated a priori using
tions (with a 1-minute rest between sets), twice a day (morn- G*Power (version 3.1.9.6)14 based on the effect size in the
ing and evening), 7 d/wk, for 4 weeks.2,3,12,31,40 VISA-A estimated from mean difference reported in earlier
RCTs.38 Considering α error = 0.05 and a study power of
Gastrocnemius, soleus, and hamstring stretch. For gastrocne- 0.95%, the sample size needed was calculated to be 24
mius stretch, the patients were instructed to stand in front of patients for each group.
the wall. With the involved leg positioned backward and the As a prerequisite for parametric analysis, data were
knee extended, the patient was instructed to lean forward screened for normality assumption using the Shapiro-Wilk
(flexing the knee of the front sound leg) while maintaining test, homogeneity of variance using Levine’s test, homoge-
the heel of the involved foot (back foot) in contact with the neity of covariance using Box’s test, and the presence of
ground. The patients were instructed to lean forward as extreme scores using outliers.
much as they could for 30 seconds. The soleus stretch was The data for functional activity (VISA-A) were normally
conducted using the same maneuver but with the involved distributed (P = .05), and homogeneity between covari-
knee slightly bent. For hamstring and gastrocnemius stretch, ances was observed (P > .05). Accordingly, a 2 × 3 mixed-
the patients were asked to lay supine on a treatment table design (generalized linear mixed model [LMM]) analysis of
with their hips of the involved limb in 90 degrees of flexion. variance (ANOVA) was used to compare the tested vari-
Then, the patients were instructed to place their hands ables of interest at different tested groups and measuring
around their involved knees (posterior aspect) and extend periods.6 However, the normality and homogeneity of vari-
them slowly until they felt a stretch at the back of the thigh ance assumptions were violated for pain as measured by the
and maintain this position for 30 seconds while pulling the VAS (raw data and after logarithmic transformation, P =
foot toward their faces. Alternatively, the hamstring and .0001). Therefore, nonparametric statistics (the Friedman
gastrocnemius were stretched from a long sitting position. test) was used to measure the differences among the 3 mea-
On a treatment table with the involved knee extended, the suring periods within each group, and Wilcoxon signed-
patients were instructed to pull their foot toward their face rank tests were used as post hoc tests wherever the Friedman
with a towel (around the ball of the foot) by both hands and test was significant. The Mann-Whitney U test was used to
sustain this position for 30 seconds. All patients in both compare both groups. The α level was set at 0.05 for all
groups were instructed to perform stretching of the gastroc- statistics. All statistical analyses were performed using
nemius, soleus, and hamstring muscles twice a day for 3 Statistical Package for the Social Sciences (SPSS version
repetitions (with 30-second stretch and 30-second rest) 7 23; SPSS, Inc).
days a week for 4 weeks. All these stretches were performed
directly after the eccentric exercises.23,34
Results
Assessments. The primary outcome measure was functional
activity of the patients as measured by the Victorian Insti-
Posttreatment Changes
tute of Sport Assessment–Achilles questionnaire (VISA- The mean values of functional activity (VISA-A score) in
A),36 which is a valid and reliable tool19 and is the gold both groups before treatment (Table 2) were comparable
standard for evaluating pain and function in AT.30 The (mixed-design ANOVA, P = .063). Mixed-design ANOVA
VISA-A consists of 8 questions covering Achilles tendon revealed significant within-group effects (F = 898.427;
pain during different situations. Questions 1 to 7 are scored P = .0001) and between-group effects (F = 181.341; P =
out of 10, while question 8 is scored out of 30. The total .0001). A significant interaction effect for intervention and
score of the VISA-A is 100. The secondary outcome mea- follow-up time was observed (F = 69.093; P = .0001).
sure was pain that was measured by the visual analog scale The functional scores improved significantly in both
(VAS), which is a well-known, valid, and reliable measur- groups after treatment (multiple pairwise post hoc com-
ing tool for pain.18 parisons, P = .0001). The improvements in the study group
Abdelkader et al 5

were better than those in the control group. The difference

(Follow-up – post)

13.6 (10.3 to 16.9)

between groups after treatment was statistically significant

−5 (8.6 to 1.4)
(multiple pairwise post hoc comparisons, P = .0001).
A significant reduction in pain (Friedman test, P =
.0001) following treatment in both groups was observed
(Table 3, expressed as median and SD). Post hoc analysis
(Wilcoxon signed-rank tests) confirmed a significant differ-
ence in VAS (P < .05) after treatment compared to that at
Mean difference (95% CI)

(Follow-up – baseline)
55.8 (51.4 to 60.2)
46.0 (41.9 to 50.1) baseline in both groups. Although both groups did not differ
regarding pain intensity at baseline (Mann-Whitney U test,
P = .867), pain reduction was significantly better in the
Table 2. VISA-A Scores and Mean Difference (as a Measure of Functional Activity) at Baseline, Posttreatment, and Follow-up for Both Groups.a

study group than in the control group (Mann-Whitney U

test, P = .0001).

Long-Term Follow-up
60.8 (56.0 to 65.5)
32.4 (28.0 to 36.9)
(Post – baseline)

At the intermediate-term follow-up (mean, 16 months),

functional scores showed a slight but still significant (P =
.003) decrease in the study group. Alternatively, the VISA-A
scores in the control group continued to increase at follow-
up (multiple pairwise post hoc comparison, P = .0001),
although they never reached the scores achieved in the
study group (Table 2). Pain scores showed a similar trend to
13 (9.8 to 16.2)
(16 months),

that in functional scores, with elevation in the study group

mean (SD)
Follow-up

and reduction in the control group at follow-up. Although

.0001
80 (5.3)b,c
67 (5.6)b,c

this change was statistically significant compared to post-

treatment (Wilcoxon signed-rank tests, P = .0001), values
Significantly different from posttreatment (multiple pairwise within-group post hoc comparison tests).

at the final follow-up were still better than those at baseline

(Table 3).
Significantly different from baseline (multiple pairwise within-group post hoc comparison tests).
31.6 (27.5 to 35.8)

At the latest (16 months) follow-up, the functional and

Posttreatment

pain scores were significantly better than those at the base-

(1 month),
mean (SD)

Abbreviation: VISA-A, Victorian Institute of Sport Assessment–Achilles questionnaire.

line (mixed-design ANOVA and Wilcoxon signed-rank

.0001
85 (6.2)b
53.4 (7.7)b

tests, respectively, P = .0001 for both). At all time points,

both scores in the study group were significantly better than
those in the control group (mixed-design ANOVA and
Mann-Whitney U test, respectively, P = .0001 for both).
P values were calculated using linear mixed-model analysis of variance.
3.3 (−0.2 to 6.7)
Baseline, mean

Between-group multiple pairwise post hoc comparison tests.

Discussion
(SD)

.063
24.2 (6.5)
21.0 (5.2)

This study is the first RCT to report the functional outcome

and pain scores for patients with noninsertional AT at an
average follow-up period of 16 months. Patients in both
groups significantly improved posttreatment (at 4 weeks)
P value (between-group comparison)d

and follow-up compared to their baseline condition. Patients

Mean difference (study vs control)

who received low-energy ESWT in addition to conven-

tional exercises had a significantly better outcome than
those who received conservative physical therapy treatment
only (eccentric training followed by stretching exercises).
These results conform to the results of studies, reviews, and
Functional activity

meta-analyses that suggested that the combination of ESWT

Control group

and exercises is an effective treatment for AT.1,13,35,38,39

Study group

One cannot miss the line of thought portrayed by RCTs

reporting the effects of ESWT in AT (Table 4). Initially,
RCTs compared shockwave therapy only to placebo,9 the
d
b
c
a
6 Foot & Ankle International 00(0)

Table 3. Descriptive Statistics and Nonparametric Tests for Visual Analog Scale (VAS) at Different Measuring Periods for Both
Groups, median (SD).a

VAS Baseline, median (SD) Posttreatment (1 month), median (SD) Follow-up (16 months), median (SD)
a
Study group 8 (1) 1 (2) 3 (2)b,c
Control group 8 (1) 7 (2)a 5 (1)b,c
P value (study vs control) 0.867 0.0001 0.0001
a
P values for comparison between groups using the Mann-Whitney U test.
b
Significantly different from baseline (using Wilcoxon signed-rank tests, post hoc).
c
Significantly different from posttreatment (using Wilcoxon signed-rank tests, post hoc).

wait-and-see policy, and eccentric exercises.39 It soon scale),38,39 indicating that the treatment was either success-
became clear that combination therapies might result in bet- ful or failed with no exact outcome scores or statistical
ter outcomes than a single modality. Rompe et al38 subse- comparison. Fan et al13 conducted a systematic review with
quently compared eccentric loading alone to a combination meta-analysis to investigate the efficacy of ESWT com-
of eccentric loading and ESWT. This combination resulted pared to other nonsurgical treatments (including sham
in a higher success rate (82%) than either eccentric loading ESWT). They reported that VAS pain scores were not sig-
or ESWT alone (60% and 52%, respectively). However, nificantly different for follow-ups shorter or longer than 6
strengthening and stretching are both essential elements in months. However, this analysis was based on 3 studies only,
the rehabilitation program for patients with AT.4 Muscle none of which was an RCT. In a study by Rompe et al,38 19
stretching after strengthening exercises has been practiced patients in group 1 (eccentric loading) and 28 patients in
by sports players for ages. Low-intensity stretching was group 2 (eccentric loading and shockwave) had a successful
shown to have moderate beneficial effects on perceived outcome based on 4-month data. Those numbers were
recovery of muscle function and postexercise muscle sore- reduced to 16 and 24, respectively, at the 1-year follow-up.
ness.5 In addition, it alleviates muscle stiffness and reduc- Despite the change in outcome scores in our study at the
tions in range of motion induced by exercise.44 Other RCTs final follow-up, all patients in this study had significantly
have therefore started to integrate stretching exercises with better scores than baseline. The latter study, as well as our
eccentric loading into their conservative (control) proto- own results, emphasizes the importance of reporting longer-
cols35,45 and compared this incorporation to the same pro- term data for studies involving treatment of AT.
gram along with ESWT. Those RCTs only reported midterm The results of this study support incorporating ESWT, in
follow-up data (3-4 months). Using our combined protocol, addition to conventional physical therapy, in any treatment
we obtained similar VISA-A scores in the study group (85 protocol for AT. Evidence in the literature so far has been
± 6.17 posttreatment and 80 ± 5.34 at final follow-up) to conflicting, with many RCTs and reviews supporting shock-
that reported by Rompe et al38 (86.5 ± 16). Although the wave therapy,1,13,38,39 while others have shown no benefit of
functional scores posttreatment in our combined exercise adding this modality to other therapeutic regimens.9,35,45
protocol group were slightly less favorable than those in the The latter studies, which failed to demonstrate a significant
aforementioned RCT, the score at the final follow-up (67 ± benefit of ESWT, have used high-energy shockwaves,16
5.59) was comparable to their eccentric loading group (73 casting some doubt about the appropriate energy level that
± 19). Our patients, however, had lower scores than their may be applied to such lesions. Moreover, 2 RCTs35,45 used
patients at baseline (24.22 ± 6.52 and 20.96 ± 5.21 vs 50.2 the American Orthopaedic Foot & Ankle Society (AOFAS)
± 11.1 and 50.6 ± 10.3 for the study and control groups, clinical rating system, an outcome score not specific for the
respectively). Moreover, the mean posttreatment difference condition being studied. While the VISA-A was developed
between groups in the present study was higher than that in as a valid, reliable, and responsive index for the severity of
the latter study (31.6 vs 13.5) with a much narrower confi- AT,36 the AOFAS score might not be the best scoring system
dence interval, which does not cross zero, unlike the former for patients with AT. It includes items such as ankle stability
study. and subtalar range of motion as well as alignment24; all are
The significant decrease in the functional scores and unrelated to AT and would not change over time with treat-
increasing pain scores in the study group of the present ment, a limitation acknowledged by the authors of one of
study at an average 16-month follow-up highlight the the aforementioned studies.35
importance of a longer-term follow-up for studies reporting Previous RCTs have compared either the follow-up
outcomes of treatments for AT. The only RCTs in literature (posttreatment) scores of the study groups9,35,45 or the
that included a “long-term” follow-up reported either com- changes in score between the baseline and follow-up.38,39
bined data for both the study and control groups9 or a global These 2 simple approaches usually yield the same estimated
“general assessment” of patients (on a 6-point Likert treatment effect, except in cases when there is imbalance in
 Table 4. Summary of Randomized Controlled Trials Studying the Effects of ESWT on Patients With Achilles Tendinopathy.a
Study Groups Sessions Patients Outcome Follow-up, mo Conclusion
9 2 b
Costa et al (2005) • ESWT (high energy, 0.2 mJ/mm ) 3 over 2 months 49 (2 groups) VAS (rest, walking sports) 3 and 12 No support for using shockwave (mainly mid-
• Placebo (bubble wrap) FIL substance). Only pain with sport improved, not
EQol-5D and health score with rest or walking
Clinical assessment Detailed outcomes only reported at 3 months
Rompe et al (2007)39 • Eccentric loading 3 over 3 weeks 75 (3 groups) VISA-A 4 Eccentric loading and ESWT comparable
• ESWT (repetitive low energy, 0.1 General assessment (Likert scale) Both better than the wait-and-see policy
mJ/mm2) Pain (NRS, pain threshold, and tenderness)
• Wait-and-see
Rasmussen et al (2008)35 • Stretching and eccentric exercise + 4 over 4 weeks 48 (2 groups) AOFAS 1, 2, and 3 ESWT group better function (more so at 8 and
rESWT (0.12-0.51 mJ/mm2) VAS (walking, stairs, working, and running) 12/52)
• Stretching and eccentric exercise No significant difference in pain
+ sham ESWT supplements treatment
Rompe et al (2009)38 • Eccentric loading 3 over 3 weeks 68 (2 groups) VISA-A, general assessment and pain 4 and 12c Combination (eccentric loading + ESWT) more
• Eccentric loading + ESWT (low assessment (see Rompe et al39) effective
energy, 0.1 mJ/mm2)
Vahdatpour et al • Conservative plan (stretching 4 over 4 weeks 43 (2 groups) AOFAS 1 and 4 ESWT group improved but not statistically
(2018)45 exercises, massage, eccentric training radial and VAS significant
and diclofenac) + ESWT focused AOFAS score only significantly different at 4 months
• Conservative plan + sham ESWT shockwaved Type of AT not specified
Rompe et al (2008)37 • Eccentric loading 3 over 3 weeks 50 (2 groups) VISA-A, general assessment, and pain 4 and 15c ESWT better in (insertional) Achilles
• ESWT (low energy, 0.12 mJ/mm2) assessment (see Rompe et al39) tendinopathy
Pinitkwamdee et al • Stretching exercise + rESWT 4 over 4 weeks 31 (2 groups) VAS 6 No difference at 6/12
(2020)33 (low energy, 0.12-0.16 mJ/mm2) VAS-FA ESWT short period of effect at 1-3/12 (insertional
• Stretching exercise + sham tendinopathy)

Abbreviations: AOFAS, American Orthopaedic Foot & Ankle Society; AT, Achilles tendinopathy; EQol-5D, EuroQol- 5 Dimension; ESWT, extracorporeal shockwave therapy; FIL, functional index of lower limb activity;
NRS, numerical rating scale; rESWT, radial extracorporeal shockwave therapy; VAS, visual analog scale; VAS-FA, visual analog scale–foot and ankle; VISA-A, Victorian Institute of Sport Assessment–Achilles questionnaire.
a
Randomized controlled trials (RCTs) reporting on insertional Achilles tendinopathy are bolded.
b
Reported only combined 1-year follow-up data for both groups. No exact figures or statistical analysis given.
c
Reported only general assessment data at 1-year follow-up, scored on a 6-point Likert scale. No pain or functional scores given and no statistical analysis.
d
High-energy ESWT (focused 0.25-0.4 mJ/mm2 + rESWT 1.8-2.6 mJ/mm2).

7
8 Foot & Ankle International 00(0)

the baseline scores of the groups.46 In fact, the only group of above MCID or otherwise, using the most commonly
researchers who calculated score change from baseline for reported MCID in clinical trials (10 points). The difference
every patient and aimed to apply the difference-in-differ- between the study and control groups in the present study at
ences model never actually reported the figures for this the 16-month follow-up (13.6 points; Table 2) exceeded
change or the treatment effect for any of the groups.38,39 both MCID reference values. The long-term benefit of
Instead, like all other studies, they only reported the mean incorporating ESWT in the treatment protocol of noninser-
group scores (at baseline and follow-up) and the mean dif- tional AT demonstrated in this study is not merely statisti-
ference between the intervention and control groups, cast- cally significant but also clinically relevant.
ing a shadow of doubt about the usefulness of such approach. This study has several strengths. In the systematic
Moreover, the difference-in-differences model (which is review by Al-Abbad et al,1 none of the studies had thera-
mainly used in economics, law, and recently to evaluate pist blinding and only 2 reported participant blinding.
changes in health care policy11,20,25) requires a preassump- Although the therapist in our study was not blinded, we
tion of parallel trends between the 2 groups prior to the managed to keep our participants blinded by using sham
intervention, an assumption that is only imperially tested at ESWT. While other studies37-39 did not blind the outcome
the best.11,20 A more robust approach that deals with such assessor, we avoided this detection bias by having an inde-
confounders is to use a linear regression model (analysis of pendent blinded assessor. This is the first study to report
covariance), which adjusts each patient’s follow-up score longer-term (mean, 16 months) detailed outcome scores of
for his or her baseline scores. Such model also has greater pain and function with complete follow-up of all patients.
statistical power to detect treatment effect.46 Alternatively, this study has some limitations. First, there
A variety of statistical approaches for comparison of was a lack of outcome data in the interim period between
means have been used in earlier RCTs (2-sample t test,9,45 posttreatment (at 4 weeks) and the final follow-up. We
Mann-Whitney U test,9 Wilcoxon test,38,39 1-way recommend keeping constant track of the outcome scores
ANOVA,38,39 and/or repeated-measures ANOVA35,45). Some throughout the whole study period, although this might
authors38,39 inappropriately analyzed data from their designs affect participant compliance. Second, the assessor in our
by collapsing across or ignoring one of the random factors study translated the VISA-A score to the patients due to
so that such familiar procedures could be used,22 which is the unavailability of a validated Arabic version of this
perhaps oversimplistic. Participants in any particular study evaluation tool. Given the fact that the assessor was
should be treated as a random factor, meaning that they are blinded, this should not have resulted in any misinterpreta-
thought to be a sample of participants that might have been tion or bias.
used.22 The mixed-model ANOVA used in this study is a
powerful analytic approach for complex research designs Conclusion
involving 2 random factors (hereafter called participants
and targets), which may be either crossed or nested, and 1 Combining calf eccentric loading with stretching exercises
fixed factor (condition).15,22 It is in essence a combination resulted in significant improvements in the pain and func-
of between-unit ANOVA and a within-unit ANOVA.15 Our tional scores in patients with noninsertional AT. Adding
study groups were “nested under” 1 categorical indepen- ESWT to this combined treatment protocol resulted in sig-
dent variable (participants) and were “crossed with” the nificantly greater improvement in both scores in the short
other independent continuous variable (time). The goal was and long term. The latter program is strongly recommended
“to determine whether the mean condition difference (treat- for treating chronic noninsertional AT.
ment outcome), given the variability of participants, is suf-
ficiently large to permit the belief that it would continue to Acknowledgments
be found with other samples of participants.”22 Mixed- We thank the Egyptian Knowledge Bank (EKB.eg) for providing
model ANOVA has substantial advantages over both tradi- expert free-of-charge copyediting/language editing through
tional and repeated-measures ANOVA, as it allows handling Enago, Crimson Interactive Inc., as part of their initiative to assist
of incomplete and unbalanced data and avoids information Egyptian researchers gain international recognition.
loss due to averaging over stimuli or participants.21
In a review of 46 studies, Murphy et al30 found evidence Declaration of Conflicting Interests
to suggest that VISA-A was the only valid and reliable self- The author(s) declared no potential conflicts of interest with
reported outcome measure of pain and function for patients respect to the research, authorship, and/or publication of this arti-
with noninsertional AT. Whereas the minimum clinically cle. ICMJE forms for all authors are available online.
important difference (MCID) established for VISA-A score
in patients with insertional AT was 6.5 points,29 the latter Funding
authors report that no clear consensus exists in case of non- The author(s) received no financial support for the research,
insertional AT.30 They suggest the utilization of either the authorship, and/or publication of this article.
Abdelkader et al 9

ORCID iDs 16. Furia JP. High-energy extracorporeal shock wave therapy as
a treatment for chronic noninsertional Achilles tendinopathy.
Nasr Awad Abdelkader, PhD, https://orcid.org/0000-0001-
Am J Sports Med. 2008;36(3):502-508.
8629-211X
17. Gerdesmeyer L, Mittermayr R, Fuerst M, et al. Current
Emad S. B. Saweeres, FRCS (Tr & Orth), MD, https://orcid
evidence of extracorporeal shock wave therapy in chronic
.org/0000-0002-6187-4305
Achilles tendinopathy. Int J Surg. 2015;24:154-159.
18. Hawker GA, Mian S, Kendzerska T, French M. Measures of
References adult pain: Visual Analog Scale for Pain (VAS Pain), Numeric
1. Al-Abbad H, Simon JV. The effectiveness of extracorporeal Rating Scale for Pain (NRS Pain), McGill Pain Questionnaire
shock wave therapy on chronic Achilles tendinopathy: a sys- (MPQ), Short-Form McGill Pain Questionnaire (SF-MPQ),
tematic review. Foot Ankle Int. 2013;34(1):33-41. Chronic Pain Grade Scale (CPGS), Short Form-36 Bodily
2. Alfredson H, Cook J. A treatment algorithm for managing Pain Scale (SF-36 BPS), and Measure of Intermittent and
Achilles tendinopathy: new treatment options. Br J Sports Constant Osteoarthritis Pain (ICOAP). Arthritis Care Res.
Med. 2007;41(4):211-216. 2011;63(S11):S240-S252.
3. Alfredson H, Pietilä T, Jonsson P, Lorentzon R. Heavy-load 19. Iversen JV, Bartels EM, Langberg H. The Victorian Institute
eccentric calf muscle training for the treatment of chronic of Sports Assessment—Achilles Questionnaire (VISA-A)—a
Achilles tendinosis. Am J Sports Med. 1998;26(3):360- reliable tool for measuring Achilles tendinopathy. Int J Sports
366. Phys Ther. 2012;7(1):76-84.
4. Allison GT, Purdam C. Eccentric loading for Achilles ten- 20. Jiménez JL, Perdiguero J. Difference-in-difference. In:
dinopathy—strengthening or stretching? Br J Sports Med. Marciano A, Ramello GB, eds. Encyclopedia of Law and
2009;43(4):276. Economics. Springer; 2019:551-555.
5. Apostolopoulos NC, Lahart IM, Plyley MJ, et al. The effects 21. Judd CM, Westfall J, Kenny DA. Treating stimuli as a ran-
of different passive static stretching intensities on recov- dom factor in social psychology: A new and comprehensive
ery from unaccustomed eccentric exercise—a randomized solution to a pervasive but largely ignored problem. J Pers
controlled trial. Appl Physiol Nutr Metab. 2018;43(8): Soc Psychol. 2012;103(1):54-69.
806-815. 22. Judd CM, Westfall J, Kenny DA. Experiments with more
6. Boisgontier MP, Cheval B. The ANOVA to mixed model than one random factor: designs, analytic models, and statisti-
transition. Neurosci Biobehav Rev. 2016;68:1004-1005. cal power. Annu Rev Psychol. 2017;68(1):601-625.
7. Chimenti RL, Cychosz CC, Hall MM, Phisitkul P. Current 23. Kedia M, Williams M, Jain L, et al. The effects of con-
concepts review update: insertional Achilles tendinopathy. ventional physical therapy and eccentric strengthening for
Foot Ankle Int. 2017;38(10):1160-1169. insertional Achilles tendinopathy. Int J Sports Phys Ther.
8. Cook JL, Stasinopoulos D, Brismée JM. Insertional and 2014;9(4):488-497.
mid-substance Achilles tendinopathies: eccentric train- 24. Kitaoka HB, Alexander IJ, Adelaar RS, Nunley JA, Myerson
ing is not for everyone—updated evidence of non-surgical MS, Sanders M. Clinical rating systems for the ankle-
management. J Manual Manipulative Ther. 2018;26(3): hindfoot, midfoot, hallux, and lesser toes. Foot Ankle Int.
119-122. 1994;15(7):349-353.
9. Costa ML, Shepstone L, Donell ST, Thomas TL. Shock 25. Lee M-J. Matching, Regression Discontinuity, Difference in
wave therapy for chronic Achilles tendon pain: a ran- Differences, and Beyond. Oxford University Press; 2016.
domized placebo-controlled trial. Clin Orthop Relat Res. 26. Longo UG, Ronga M, Maffulli N. Achilles tendinopathy.
2005;440:199-204. Sports Med Arthrosc Rev. 2009;17(2):112-126.
10. Courville XF, Coe MP, Hecht PJ. Current concepts review: 27. Maffulli N, Longo UG, Kadakia A, Spiezia F. Achilles tendi-
noninsertional Achilles tendinopathy. Foot Ankle Int. nopathy. Foot Ankle Surg. 2020;26(3):240-249.
2009;30(11):1132-42. 28. Mafi N, Lorentzon R, Alfredson H. Superior short-term
11. Dimick JB, Ryan AM. Methods for evaluating changes in results with eccentric calf muscle training compared to con-
health care policy: the difference-in-differences approach. centric training in a randomized prospective multicenter
JAMA. 2014;312(22):2401-2402. study on patients with chronic Achilles tendinosis. Knee Surg
12. Fahlström M, Jonsson P, Lorentzon R, Alfredson H. Chronic Sports Traumatol Arthrosc. 2001;9(1):42-47.
Achilles tendon pain treated with eccentric calf-muscle train- 29. McCormack J, Underwood F, Slaven E, Cappaert T. The
ing. Knee Surg Sports Traumatol Arthrosc. 2003;11(5): minimum clinically important difference on the VISA-A and
327-333. LEFS for patients with insertional Achilles tendinopathy. Int
13. Fan Y, Feng Z, Cao J, Fu W. Efficacy of Extracorporeal J Sports Phys Ther. 2015;10(5):639-644.
Shock Wave Therapy for Achilles Tendinopathy: A Meta- 30. Murphy M, Rio E, Debenham J, Docking S, Travers M,
analysis. Sage; 2020. Gibson W. Evaluating the progress of mid-portion Achilles
14. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a tendinopathy during rehabilitation: a review of outcome mea-
flexible statistical power analysis program for the social, sures for self- reported pain and function. Int J Sports Phys
behavioral, and biomedical sciences. Behav Res Methods. Ther. 2018;13(2):283-292.
2007;39(2):175-191. 31. Öhberg L, Lorentzon R, Alfredson H. Eccentric training in
15. Frey BB. The SAGE Encyclopedia of Educational Research, patients with chronic Achilles tendinosis: normalised tendon
Measurement, and Evaluation. Sage; 2018.
10 Foot & Ankle International 00(0)

structure and decreased thickness at follow up. Br J Sports of the main body of tendo Achillis: a randomized controlled
Med. 2004;38(1):8-11. trial. Am J Sports Med. 2007;35(3):374-383.
32. Pearce CJ, Tan A. Non-insertional Achilles tendinopathy. 40. Sayana MK, Maffulli N. Eccentric calf muscle training in
EFORT Open Rev. 2016;1(11):383-390. non-athletic patients with Achilles tendinopathy. J Sci Med
33. Pinitkwamdee S, Laohajaroensombat S, Orapin J, Woratanarat Sport. 2007;10(1):52-58.
P. Effectiveness of extracorporeal shockwave therapy in the 41. Silbernagel KG, Hanlon S, Sprague A. Current clinical con-
treatment of chronic insertional Achilles tendinopathy. Foot cepts: conservative management of Achilles tendinopathy. J
Ankle Int. 2020;41(4):403-410. Athletic Train. 2020;55(5):438-447.
34. Porter D, Barrill E, Oneacre K, May BD. The effects of dura- 42. Stania M, Juras G, Chmielewska D, Polak A, Kucio C, Król
tion and frequency of Achilles tendon stretching on dorsiflex- P. Extracorporeal shock wave therapy for Achilles tendinopa-
ion and outcome in painful heel syndrome: a randomized, thy. BioMed Res Int. 2019;2019:3086910.
blinded, control study. Foot Ankle Int. 2002;23(7):619-624. 43. Sussmilch-Leitch SP, Collins NJ, Bialocerkowski AE,
35. Rasmussen S, Christensen M, Mathiesen I, Simonson O. Warden SJ, Crossley KM. Physical therapies for Achilles
Shockwave therapy for chronic Achilles tendinopathy: a tendinopathy: systematic review and meta-analysis. J Foot
double-blind, randomized clinical trial of efficacy. Acta Ankle Res. 2012;5(1):15.
Orthopaedica. 2008;79(2):249-256. 44. Torres R, Appell HJ, Duarte JA. Acute effects of stretching
36. Robinson JM, Cook JL, Purdam C, et al. The VISA-A ques- on muscle stiffness after a bout of exhaustive eccentric exer-
tionnaire: a valid and reliable index of the clinical severity of cise. Int J Sports Med. 2007;28(7):590-594.
Achilles tendinopathy. Br J Sports Med. 2001;35(5):335-341. 45. Vahdatpour B, Forouzan H, Momeni F, Ahmadi M, Taheri
37. Rompe JD, Furia J, Maffulli N. Eccentric loading compared P. Effectiveness of extracorporeal shockwave therapy for
with shock wave treatment for chronic insertional Achilles chronic Achilles tendinopathy: a randomized clinical trial. J
tendinopathy. a randomized, controlled trial. J Bone Joint Res Med Sci. 2018;23:37.
Surg. 2008;90(1):52. 46. Vickers AJ, Altman DG. Statistics notes: analysing con-
38. Rompe JD, Furia J, Maffulli N. Eccentric loading versus trolled trials with baseline and follow up measurements. BMJ.
eccentric loading plus shock-wave treatment for midportion 2001;323(7321):1123-1124.
Achilles tendinopathy: a randomized controlled trial. Am J 47. Wilson F, Walshe M, O’Dwyer T, Bennett K, Mockler D,
Sports Med. 2009;37(3):463-470. Bleakley C. Exercise, orthoses and splinting for treating
39. Rompe JD, Nafe B, Furia JP, Maffulli N. Eccentric loading, Achilles tendinopathy: a systematic review with meta-analy-
shock-wave treatment, or a wait- and-see policy for tendinopathy sis. Br J Sports Med. 2018;52(24):1564-1574.