A comparative analysis of predictive models of morbidity in intensive care unit after cardiac surgery – Part II: an illustrative example

Table 7 Variables entered and removed (in square brackets) at each step of the stepwise selection procedure

Step no.	BL	BQ	k NN	LR	HS	DS	ANN1	ANN2
1	O₂ER	O₂ER	Post-CI	O₂ER	O₂ER	SvO₂	O₂ER	O₂ER
2	VCO₂	DO₂I	O₂ER	VCO₂	VCO₂	Card-ID	Card-ID	VO₂
3	Card-ID	Card-ID	Card-ID	Card-ID	Card-ID	DO₂I	VO₂	Card-ID
4	PVD	PVD	PVD	PVD	PVD	O₂ER	PVD	PVD
5	TBU	W	TBU	TBU	TBU	PVD	TBU	Gly
6	EM	VD	MAP	EM	EM	O₂ER	Pre-CI	Gender
7	Pre-CI	DAP		SAP	SAP	EM	EM	MVR
8	WBC	SAP		Pre-CI	Pre-CI	BSA	WBC	Cr
9	SAP	Diur		WBC	WBC	AD	Age	AVO₂
10	Age	Xclamp		SaO₂	SaO₂	CHF	SaO₂	Arrhy
11	PvO₂	H		PvO₂	PvO₂	MVR	AD
12	SaO₂	Gender		AD	AD	O₂ER	P/F
13	AD			PaO₂	PaO₂	MR	PVS
14	SvO₂			Cr	Cr	EM	Temp
15	Xclamp			CvO₂	CvO₂	Card-ID
16	PaO₂			Xclamp	Xclamp	O₂ER
17	[PvO₂]			DO₂I	DO₂I	PVD
18	Cr			W	W	Diab
19	Intra-CI			SVRI	SVRI	VCO₂
20	Post-CI			[VCO₂]	[VCO₂]	O₂ER
21	[Age]			Pre-IABP	Pre-IABP	AD
22	W			CHF	CHF	O₂ER
23	[Pre-CI]			[Xclamp]	[Xclamp]	CHF
24	DO₂I					CABG-C
25	[VCO₂]					MR
26	Bil					PAH
27	Hb					O₂ER
28	[WBC]					IABP
29	SVRI					O₂ER
30	[Post-CI]					IABP
31	BSA					O₂ER
32	CABG-A					PVS
33						IABP
34						O₂ER
35						IABP

BL, Bayes linear model; BQ, Bayes quadratic model; k NN, k-nearest neighbour model; LR, logistic regression model; HS, Higgins score system; DS, direct score system; ANN1, one-layer artificial neural network; ANN2, two-layer artificial neural network. Predictor variable abbreviations are indicated in Tables 1-6.

ISSN: 1472-6947