204
A. Morral et al.
Outcome measures
The primary outcome was foot function, measured using the
FFI, a self-administered questionnaire that quantifies the impact
of pathology on foot function in terms of pain, disability and
activity restriction (24), and is one of the most-used assessment
tools in clinical trials on plantar fasciitis treatments (25, 26).
The index consists of 23 items that assess foot function on a
continuous scale of 100 points, where a lower score indicates
better function. This variable was assessed at baseline and at
4 follow-up sessions: at 1, 2, 4 and 14 months after the last
rESWT session.
Secondary variables included: (i) pain assessment for “pain
with the first weight-bearing step in the morning” and “pain
during the day”, measured with a 10-point VAS at baseline and
at 4 follow-ups: at 1, 2, 4 and 14 months after the last rESWT
session; (ii) plantar fascia thickness of the affected foot, mea-
sured using ultrasound (Echo Blaster EXT-128) at baseline and
at 2 time-points: 4 and 14 months after the last rESWT session.
Complications, adverse effects during and after treatment were
recorded. Information was also collected on discomfort during
the application of rESWT (using a VAS scale).
Baseline patient data collected included age, sex, body mass
index (BMI) (kg/m 2 ), duration of symptoms (months), time
spent standing per day (h), exercise level (1, no activity; 2, less
than 5 h a week; 3, between 5 and 10 h a week; 4, more than 10
h a week), and level of education (1, no education; 2, compul-
sory education (primary and secondary school); 3, high-school
studies or vocational training; and 4, university-level education).
Randomization and blinding
Randomization was performed through a specific syntax for this
study, in IBM-SPSS language (V22), in blocks and perfectly
balanced by a statistician from the Ibero-American Cochrane
Centre, Clinical Epidemiology and Public Health Service of the
Hospital de la Santa Creu i Sant Pau. The randomization num-
bers were placed in sealed opaque envelopes, thus concealing
allocation from both the patients and the therapist until treatment
started. The envelopes were sequentially ordered from 1 to 135
and stored in a locked cabinet. Both the lead researcher (who
assessed patients for eligibility) and the researcher responsible
for delivering the treatment had no access to the content of the
envelopes, ensuring the concealment of assignments. The pa-
tients and the therapist were not blinded to the treatment due to
the nature of the study, as they knew which of the 3 devices was
being applied, but the evaluator and data analyst were blinded.
Statistical analysis
Descriptive values are expressed as means and standard devia-
tions (SD) for quantitative variables, and relative frequency for
categorical variables.
As the main study analysis, a 2-way analysis of variance
(ANOVA) was performed, with groups (3 levels: (groups I, II
and III)) and time (5 levels (baseline and follow-ups) of repeated
measures) as independent variables, and clinical variables (foot
functionality, morning pain, pain during the day, and plantar
fascia thickness) as dependent variables.
For secondary analysis with categorical variables a χ 2 test
was performed. Finally, in the case of quantitative variables,
the inferential analysis was carried out with analysis of va-
riance (ANOVA); the Kruskal–Wallis test was performed as
a non-parametric test alternative to ANOVA (failure to meet
assumptions or for ordinal variables).
The sample size was calculated using GRANMO Software
(Version 7.12, IMIM-Hospital del Mar, Barcelona, Spain), with
the assumption that an ANOVA would be performed among the
3 groups and 5 time-points. Accepting a p-value below 0.05 and
a statistical power (beta-1 risk) above 0.8 in a bilateral contrast,
45 patients were required in each group to detect a minimum
difference of 15 points between 2 groups, assuming that there
were 3 groups and a standard deviation (SD) of 20. A maximum
rate of 15% loss to follow-up was estimated.
Calculations were performed using IBM-SPSS (Version 22,
IBM, Armonk, NY, USA).
RESULTS
Baseline patient data
A total of 135 patients participated in the study; 45 in
each group. Mean age was 49 years, and 48% of the
patients were women. Mean duration of symptoms was
14 months. Table II shows the baseline patient data.
Table II. Baseline data
Age, years
Body mass index, kg/m 2
Duration of symptoms, months
Hours spent standing per day
Sex, n (%)
Female
Male
Exercise, n (%)
No activity
Low activity
Moderate activity
Intense activity
Educational level, n (%)
No education
Compulsory education
High-school education or vocational training
University education
Group I: standard device
(n = 45)
Mean (SD) Group II: sophisticated device
(n = 45)
Mean (SD) Group III: austere
device (n = 45)
Mean (SD) p-value
48.27 (9.96) 52.51 (12.28) 49.31 (11.14) 0.054*
27.61 (3.31)
13.27 (9.26)
7.67 (3.53) 27.66 (3.72)
14.98 (12.72)
8.80 (3.67) 28.35 (5.25)
13.53 (9.09)
8.33 (4.09) 0.642
0.708
0.362
33.3
66.7 60
40 51.1
48.9 0.036*
44.4
35.6
15.6
4.4 44.4
42.2
13.3
0 57.8
28.9
8.6
4.4 0.407
2.2
35.6
33.3
28.9 8.9
44.4
28.9
17.8 6.7
31.1
33.3
28.9
0.166
*At baseline, significant differences were observed between groups for the distribution of the variable sex (p = 0.036), and for the variable age, the difference
was close to significance (p = 0.054). Thus, the entire analysis was corrected for the sex and age of the patients, using a co-variance analysis (ANCOVA). SD:
standard deviation.
www.medicaljournals.se/jrm