35 / 66
CARDIOVASCULAR JOURNAL OF AFRICA • Volume 30, No 6, November/December 2019
AFRICA
343
for three different walking speeds: slow (illustrated by the turtle),
normal (illustrated by a human) and fast (illustrated by a rabbit).
If unable to do the task (e.g. walk fast), they were instructed to
put a cross on the watch screen. A fourth and last item requires
the patient to estimate his/her usual pace: very slow (illustrated
by a snail), slow (illustrated by a turtle), normal (illustrated by
a human) or fast (illustrated by a rabbit). Illustrations of the
WELSH tool are presented in Fig. 1.
Patients were provided a blue or black pen and reading glasses
if needed, and received oral explanation of the protocol and on
how to fill in the WELSH questionnaire in their native language
or dialect. Patients were asked to self-complete each of the three
watch dials and define whether they felt they walked at their
usual pace. The two series of drawings (first three items about
maximal duration at different paces and fourth item on usual
pace) were printed on each side of the same sheet and were
thereby submitted in a random order.
Once self-completed, the technician, nurse or physician checked
the questionnaire and completed any incomplete questionnaires.
For the WELSH tool, an error was defined as a missing answer,
multiple answers to the same item or a reported increase in the
duration of a task that was more difficult than the task of lower
intensity (e.g. reported higher ability to walk fast than to walk
at a normal speed). Errors were discussed with the patient and
corrections or additions were made and written in red.
We empirically defined a method for scoring of the
questionnaire that could easily be memorised and done by
mental calculation and would result in a maximal score of 100.
The number of points attributed to the possible durations ranged
from zero points if the task was impossible, to eight points for
51 to 60 minutes. The number of points increased by one point
for each interval of five minutes up to 20 minutes, and for each
interval of 10 minutes from 20 minutes to one hour. Estimated
usual walking speeds were attributed coefficients ranging from
one for ‘much slower – snail’ to four for ‘faster – rabbit’. The
final score was the sum of the number of points obtained at each
pace plus one, multiplied by the coefficient resulting from the
estimation of walking speed. Therefore the minimum possible
WELSH score is 1 = [(0
+
0
+
0)
+
1]
×
1, and the maximal score
is 100 = [(8
+
8
+
8)
+
1]
×
4.
All patients had a six-minute walking test to objectively
quantify their walking ability. This test was performed by
a different physician, technician or nurse from the one who
supervised the WELSH completion and blinded to the results
of the WELSH score. The test was performed between two
plastic cones positioned 30 m from each other in a corridor. Oral
instruction in the patient’s language or dialect was to: (1) walk
(never run) back and forth as many times as possible, turning
around the plastic cones; (2) cover as much distance as possible
over six minutes; (3) slow down or stop if needed, and restart
walking when possible. For all tests, we recorded the maximal
walking distance performed at minute six (MWD).
The study was approved by the Institutional Review Board
of the Ministry of Health of Burkina Faso in August 2017, and
registered on ClinicalTrials.gov Identifier: NCT03482869. It was
performed according to the international ethics standards and
conforms to the Helsinki Declaration. A signature confirming
informed consent to participate in the study was obtained from
all patients after oral (and written when possible) explanation
of the study.
Statistical analyses
We analysed the number of incomplete questionnaires to estimate
the questionnaire feasibility. The correlations of the WELSH
scores with MWD were performed with step-by-step linear
regression analysis to determine the factors associated with
MWD. Then the Pearson
r
-coefficient of correlation between the
WELSH score and MWD was calculated. Results are presented
as mean
±
standard deviation or number and percentages.
From previous studies, we estimated the
r
-coefficients of
correlation between the objective measure of walking distance
and the questionnaire to range from 0.35 to 0.66.
6,16-18
Assuming a
coefficient of correlation between WELSH and MWD of at least
0.40, the minimum number of subjects to reach a power of 80%
for two-tailed alpha equal to 0.05 was 47 subjects.
For organisational reasons, the protocol was scheduled over
a series of periods starting in March 2018, in order to recruit 50
patients. Correlation from 0.40 to 0.59 was considered fair, from
0.60 to 0.75 was considered good, and above 0.75 was considered
very good. Statistical analyses were performed with SPSS V15.0.
For all tests, a two-tailed
p
<
0.05 was used to indicate statistical
significance.
Results
Of the 50 patients, 29 (58.0%) were referred for the follow up
of treatment for chronic hypertension, 10 (20.0%) were referred
for apparent non-ischaemic cardiomyopathy, three (6.0%) had
a clinical history of infarction and four patients (8.0%) were
assumed to suffer from chronic valvular disease. Among all these
patients, only four had an echocardiography and two had chest
X-ray imaging to estimate the cardiothoracic index.
The characteristics of the 50 included patients are presented
in Table 1. Notably, two-thirds of the patients stopped their
education at elementary school. Among the 50 patients, one
was unable to complete the WELSH questionnaire even under
supervision. Among the other 49 patients, 33 patients completed
Table 1. Characteristics of study patients
Characteristics
Number (%)
Age (years)
54.8
±
10.7
Females
40 (80.0)
Residential area
Urban
47 (94)
Suburban
1 (2)
Rural
2 (4)
School level
Never been to school
19 (38.0)
Elementary
15 (30.0)
Primary
13 (26.0)
Secondary
3 (6.0)
Weight (kg)
74.9
±
15.2
Height (cm)
163
±
10
Body mass index (kg/m²)
28.6
±
9.2
Waist circumference (cm)
92.5
±
11.8
Self-reported walking limitation
33 (66.0%)
Heart failure
34 (68.0%)
New York Heart Association classification
No dyspnoea
18 (36)
I
13 (23)
II
15 (30)
III
4 (8)