Validity and reliability of a spatiotemporal gait-analysis system
Statistical analysis
Concurrent validity was evaluated from the intraclass correlation
coefficients model 2,1 (ICCs 2,1 ) using a 95% confidence interval
(95% CI), and the standard error of measurement (SEM) and per-
centages (%SEM) were calculated from an analysis of variance.
The SEM was considered small if it represented less than 5% of
the weighted mean. Systematic differences between the systems
were determined using paired t-tests. Based on the resolution of
the GAITRite® system for spatial (1.27 cm) and temporal (0.02 s)
parameters, differences smaller than these values were considered
as measurement error. Bland-Altman repeatability coefficients
(RCs) were calculated as 1.96 times the standard deviation of the
difference between the 2 systems under comparison. The RC was
considered small if it represented less than 8% of the weighted
mean, which was the RC found between the GAITRite® system
and 3-dimensional gait analysis systems (32).
To determine the minimum number of footsteps needed to
achieve an adequate level of reliability for the SGAS data,
ICCs model 3,1 (ICCs 3,1 ) were calculated per gait parameter
for 2 steps and for each incremental step (n) up to 10 steps.
Subsequently, a coefficient of variation (CoV) was calculated
between the 95% limits of agreement interval (calculated as
the mean difference of n steps and 10 steps ± 1.96 times the SD
of the difference between n steps and 10 steps) divided by the
mean value for 10 steps. Data were considered reliable when
this CoV was < 10%. Inter-rater and intra-rater reliability were
459
examined using ICCs (inter-rater: ICC 2,1 and intra-rater: ICC 3,1 )
with 95% CIs.
The following classification for the ICC was used: poor
(< 0.50), moderate (0.50–0.74), good (0.75–0.89), and excellent
(≥ 0.90). An ICC with a value of 0.90 or greater and a lower limit
of the 95% CI of at least 0.75 were considered as acceptable
(33). For all ICCs, the absolute agreement criterion was used.
The presence of heteroscedasticity was examined through vi-
sual inspection of the Bland-Altman plots. To ensure statistical
power, a sample size of at least 30 subjects was required (34).
Significance was set at p < 0.05. Data analysis was performed
using IBM SPSS Statistics 23.0 (SPSS Inc., Chicago, IL, USA).
RESULTS
Approximately 78–98% of the collected data across
walking conditions for the stationary and moving
setup of the SGAS and the GAITRite® system was
applicable. Most missing data came from invalid tri-
als with the GAITRite® system due to steps outside
the active sensor area and errors in the foot detection
of the sensors. In total, between 256 and 322 valid
trials of footsteps across conditions for the stationary
and moving setup were collected from the SGAS and
Table II. Concurrent validity and measurement error between the spatiotemporal gait analysis system (SGAS) and GAITRite® system
for all conditions for 10 footsteps per subject
SGAS
Mean (SD)
Mean difference ICC 2.1
(SD)
(95% CI) SEM %SEM RC %RC
67.58 (7.22)
0.57 (0.05)
0.70 (0.08)
0.18 (0.07) –0.98 (1.10)*
0.00 (0.01)
0.01 (0.02)*
–0.05 (0.07)* 0.98
0.99
0.97
0.21 (0.89–0.99)
(0.98–0.99)
(0.95–0.98)
(0.02–0.37) 1.05
0.01
0.01
0.05 1.6
1.1
2.0
34.5 2.16
0.02
0.04
0.13 3.2
3.0
5.1
83.2
70.33 (7.16)
0.56 (0.05)
0.69 (0.08)
0.14 (0.03) 71.57 (7.28)
0.56 (0.05)
0.69 (0.08)
0.13 (0.03) –1.24 (1.28)*
0.00 (0.01)*
0.00 (0.01)*
0.01 (0.02)* 0.97
0.97
0.98
0.81 (0.81–0.99)
(0.96–0.98)
(0.98–0.99)
(0.75–0.86) 1.26
0.01
0.01
0.01 1.8
1.7
1.5
9.3 2.50
0.03
0.03
0.03 3.5
4.6
4.1
24.8
41.21 (8.76)
2.05 (1.35)
3.12 (2.29)
1.15 (1.05)
0.90 (0.41)
78.12 (13.60)
4.19 (2.49) 41.87 (8.99)
2.05 (1.35)
3.19 (2.35)
0.87 (0.80)
0.99 (0.47)
79.36 (14.02)
4.20 (2.49) –0.67 (0.90)*
–0.01 (0.05)
0.08 (0.08)*
0.28 (0.49)*
–0.09 (0.09)*
–1.23 (1.26)*
–0.01 (0.05)* 0.99
1.00
1.00
0.83
0.96
0.99
1.00 (0.97–1.00)
(1.00–1.00)
(0.99–1.00)
(0.66–0.90)
(0.78–0.99)
(0.94–1.00)
(1.00–1.00) 0.79
0.04
0.07
0.40
0.09
1.24
0 1.9
2.1
2.3
39.1
9.4
1.6
0 1.76
0.10
0.16
0.97
0.18
2.46
0.10 4.2
5.0
5.0
95.5
19.2
3.1
2.4
43.27 (11.62)
0.64 (0.18)
0.83 (0.27)
0.36 (0.16)
0.44 (0.08)
81.87 (19.04)
1.28 (0.36) –0.46 (1.39)*
–0.00 (.02)
0.01 (0.02)*
–0.17 (0.06)*
–0.01 (0.02)*
–1.01 (1.33)*
–0.00 (0.01) 0.99
1.00
1.00
0.50
0.95
1.00
1.00 (0.99–0.99)
(1.00–1.00)
(1.00–1.00)
(–0.06–0.82)
(0.93–0.96)
(0.99–1.00)
(1.00–1.00) 1.04
0.01
0.02
0.11
0.02
1.20
0.01 2.4
1.5
1.8
40.4
3.9
1.5
0.9 2.72
0.03
0.04
0.12
0.05
2.61
0.03 6.3
4.5
4.9
44.1
10.6
3.2
2.1
69.94 (6.53)
0.53 (0.04)
0.65 (0.06)
0.12 (.02)
0.41 (.03)
139.84 (12.75)
1.05 (0.08) –0.06 (0.85)
–0.00 (0.01)
0.01 (0.01)*
0.01 (.02)*
–0.01 (.01)*
–0.10 (1.25)
–0.00 (0.01) 0.99
0.97
0.98
0.72
0.88
1.00
0.99 (0.99–0.99)
(0.96–0.98)
(0.95–0.99)
(0.55–0.82)
(0.74–0.93)
(0.98–1.00)
(0.99–0.99) 0.59
0.01
0.01
0.01
0.01
0.90
0.08 0.8
1.4
1.5
10.2
2.5
0.7
0.7 1.67
0.02
0.02
0.03
0.02
2.45
0.02 2.4
4.0
3.6
25.6
6.1
1.8
1.9
Spatiotemporal gait parameter n Comfortable gait speed a
Step length, cm
Step time, s
Stance time, s
Double support time, s 315
315
315
314 66.60 (7.27)
0.57 (0.05)
0.69 (0.07)
0.13 (0.03)
322
322
322
322 290
289
289
288
210
210
210
Shod walking a
Step length, cm
Step time, s
Stance time, s
Double support time, s
Slow gait speed a
Step length, cm
Step time, s
Stance time, s
Double support time, s
Swing time, s
Stride length, cm
Stride time, s
Toe walking a
Step length, cm
276
42.81 (11.55)
Step time, s
275
0.64 (0.18)
Stance time, s
274
0.83 (0.26)
Double support time, s
274
0.19 (0.10)
Swing time, s
151
0.43 (0.07)
Stride length, cm
150
80.88 (18.96)
Stride time, s
151
1.28 (0.36)
Comfortable gait speed by the moving SGAS camera
Step length, cm
256
69.88 (6.56)
Step time, s
259
0.52 (0.04)
Stance time, s
258
0.64 (0.06)
Double support time, s
258
0.13 (0.03)
Swing time, s
256
0.40 (0.03)
Stride length, cm
256
139.74 (12.76)
Stride time, s
256
1.05 (0.08)
GAITRite® system
Mean (SD)
*p<0.05. a Stationary SGAS camera. n: number of trials (depending on valid trials of the SGAS and GAITRite® system for all 33 subjects); SD: standard deviation;
ICC: intraclass correlation coefficient; CI: confidence interval; SEM: standard error of measurement; %SEM: standard error of measurement as percentage of
the weighted mean; RC: repeatability coefficient; %RC: repeatability coefficient as percentage of the weighted mean.
J Rehabil Med 51, 2019