Regression models for analyzing radiological visual grading studies – an empirical comparison

BMC Medical Imaging

Table 1 Estimated parameters, goodness-of-fit statistics and estimated dose reduction for GWscore

Model	Coefficient						Goodness-of-fit		Dose reduction
Model	logCTDI		id2		id4		AIC	Pseudo R²	id2	id4
	Est.	P-value	Est.	P-value	Est.	P-value	AIC	Pseudo R²	id2	id4
regress^a	1.459	<0.001	0.158	<0.001	0.208	<0.001	-	0.4160	10.29 %	13.31 %
regress^a	(1.244, 1.674)		(0.082, 0.234)		(0.132, 0.284)		-	0.4160	(6.14 %, 14.43 %)	(9.37 %, 17.24 %)
ologit^a	8.825	<0.001	0.966	<0.001	1.271	<0.001	1124.35	0.4172	10.37 %	13.41 %
ologit^a	(7.354, 10.295)		(0.512, 1.419)		(0.812, 1.730)		1124.35	0.4172	(6.35 %, 14.39 %)	(9.60 %, 17.23 %)
gologit2^a =2	9.487	<0.001	1.262	<0.001	1.465	<0.001	1184.56	0.4342	12.45 %	14.31 %
gologit2^a =2	(7.213, 11.761)		(0.682, 1.842)		(0.873, 2.058)				(7.44 %, 17.46 %)	(9.25 %, 19.37 %)
gologit2^a =3	8.165	<0.001	0.521	0.172	0.985	0.008			6.18 %	11.37 %
gologit2^a =3	(6.143, 10.189)		(−.227, 1.269)		(0.260, 1.711)				(−1.62 %, 13.98 %)	(4.87 %, 17.86 %)
slogit^a	17.447	<0.001	1.887	<0.001	2.433	<0.001	1123.27	0.4201	10.25 %	13.05 %
slogit^a	(14.460, 20.435)		(1.028, 2.746)		(1.555, 3.310)		1123.27	0.4201	(6.35 %, 14.15 %)	(9.23 %, 16.80 %)
mixed^b	1.459	<0.001	0.158	<0.001	0.208	<0.001	1225.30	0.2748	10.29 %	13.31 %
mixed^b	(1.244, 1.673)		(0.082, 0.234)		(0.132, 0.284)		1225.30	0.2748	(6.14 %, 14.43 %)	(9.38 %, 17.24 %)
meologit^b	8.433	<0.001	0.922	<0.001	1.213	<0.001	1215.96	0.2751	10.36 %	13.4 %
meologit^b	(6.685, 10.180)		(0.452, 1.392)		(0.735, 1.692)		1215.96	0.2751	(6.21 %, 14.51 %)	(9.49 %, 17.32 %)

ISSN: 1471-2342