Noninferiority and Equivalence Evaluation of Clinical Performance among Computed Radiography, Film, and Digitized Film for Telemammography Services
Table 4
Equivalence tests () for TN, TP, FN, and FP by paired devices.
Variable
Compared devices
Mean difference ()
SE
(1-2α)% confidence interval for equivalence testing
() device
() device
Lower
Upper
TN
LUMIX
ICR
−0.006
0.0102
−0.023
0.011
−9.22
<0.001
Ha
LUMIX
CR
−0.006
0.0084
−0.020
0.008
−11.20
<0.001
Ha
LUMIX
FILM
−0.016
0.0098
−0.032
0.000
−8.53
<0.001
Ha
ICR
CR
0.000
0.0091
−0.015
0.015
−10.96
<0.001
Ha
ICR
FILM
−0.010
0.0097
−0.026
0.006
−9.25
<0.001
Ha
FILM
CR
0.010
0.0067
−0.001
0.021
−13.49
<0.001
Ha
TP
LUMIX
ICR
0.016
0.0098
0.000
0.032
−8.53
<0.001
Ha
LUMIX
CR
−0.008
0.0104
−0.025
0.009
−8.86
<0.001
Ha
LUMIX
FILM
−0.031
0.0122
−0.051
−0.011
−5.68
<0.001
Ha
ICR
CR
−0.024
0.0104
−0.042
−0.007
−7.26
<0.001
Ha
ICR
FILM
−0.047
0.0135
−0.069
−0.025
−3.94
<0.001
Ha
FILM
CR
0.022
0.0152
−0.003
0.047
−5.11
<0.001
Ha
FN
LUMIX
ICR
−0.016
0.0098
−0.032
0.000
−8.53
<0.001
Ha
LUMIX
CR
0.008
0.0104
−0.009
0.025
−8.86
<0.001
Ha
LUMIX
FILM
0.031
0.0122
0.011
0.051
−5.68
<0.001
Ha
ICR
CR
0.024
0.0104
0.007
0.042
−7.26
<0.001
Ha
ICR
FILM
0.047
0.0135
0.025
0.069
−3.94
<0.001
Ha
FILM
CR
−0.022
0.0152
−0.047
0.003
−5.11
<0.001
Ha
FP
LUMIX
ICR
0.006
0.0102
−0.011
0.023
−9.22
<0.001
Ha
LUMIX
CR
0.006
0.0084
−0.008
0.020
−11.20
<0.001
Ha
LUMIX
FILM
0.016
0.0098
0.000
0.032
−8.53
<0.001
Ha
ICR
CR
0.000
0.0091
−0.015
0.015
−10.96
<0.001
Ha
ICR
FILM
0.010
0.0097
−0.006
0.026
−9.25
<0.001
Ha
FILM
CR
−0.010
0.0067
−0.021
0.001
−13.49
<0.001
Ha
Each comparison was calculated from 980 observations (70 cases, seven radiologists, and two devices). TN: true negative proportion, TP: true positive proportion, FN: false negative proportion, FP: false positive proportion, SE: standard error of the mean, Ho: null hypothesis, Ha: alternative hypothesis for testing equivalence, α: significance of the test (0.05), δ: difference of the means allowed to achieve equivalence, : test for difference of compared devices, that is, = (Difference () − δ)/SE, : retained hypothesis equivalence at level (“Ha” indicates equivalence achieved and “Ho” indicates failing to reject the null hypothes). Ho: difference () − δ = 0. Ha: difference () − δ < 0.