Table 4.

Logistic regression analyses for hospital mortality of elderly group.

Parameters	β-coefficient	Standard error	Odds ratio, 95% CI	P-value	Adjusted for sparse data bias
Univariate logistic regression
Female	1.946	0.816	7.00 (1.41–34.68)	0.017
ESRD	2.449	1.230	11.57 (1.04–128.97)	0.047
Emergency surgery	1.792	0.832	6.00 (1.18–30.62)	0.031
Preoperative inotropic support	2.449	1.230	11.57 (1.04–128.97)	0.047
Preoperative ICU hospitalization	2.120	0.879	8.33 (1.49–46.71)	0.016
LVEF	0.092	0.045	1.10 (1.00–1.20)	0.042
Multivariate logistic regressiona
Female	2.205	1.357	9.07 (0.64–129.55)	0.104	Before
	1.493	0.570	4.45 (1.46–13.72)	0.009	After
ESRD	4.620	1.813	104.49 (2.91–3544.27)	0.011	Before
	1.949	0.622	7.02 (2.07–23.85)	0.002	After
Emergency surgery	1.640	1.662	5.15 (0.20–133.81)	0.774	Before
	1.316	0.594	3.73 (1.16–11.93)	0.027	After
Preoperative inotropic support	–0.024	2.345	0.98 (0.01–96.69)	0.992	Before
	0.060	0.642	1.06 (0.30–3.76)	0.926	After
Preoperative ICU hospitalization	2.625	1.778	13.80 (0.42–450.19)	0.140	Before
	1.506	0.604	4.51 (1.38–14.74)	0.013	After
LVEF	0.019	0.066	1.02 (0.90–1.16)	0.774	Before
	0.026	0.047	1.03 (0.94–1.14)	0.575	After

^aHosmer-Lemeshow test, P = 0.844.

ESRD, end-stage renal disease; ICU, intensive care unit; LVEF, left ventricular ejection fraction.