Table 5.

Bivariate and multivariable analysis of the association between natremia levels and in-hospital or 30-day mortality: Lausanne University Hospital, 2012–2018.

	In-hospital mortality					30-day mortality
	Bivariate			Multivariable		Bivariate			Multivariable
	Alive	Dead	p value	HR (95% CI)	p value	Alive	Dead	p value	HR (95% CI)	p value
Natremia levels			<0.001
Normal	2,560 (68.2)	138 (50.2)		1 (reference)		2,473 (68.5)	225 (53.2)	<0.001	1 (reference)
Decreased	1,196 (31.8)	137 (49.8)		1.82 (1.43–2.33)	<0.001	1,135 (31.5)	198 (46.8)		1.68 (1.37–2.04)	<0.001
Natremia levels			<0.001
Normal	2,560 (68.2)	138 (50.2)		1 (reference)		2,473 (68.5)	225 (53.2)	<0.001	1 (reference)
Mild (130–135 mEq/L)	825 (22)	78 (28.4)		1.48 (1.12–1.97)	0.007	792 (22.0)	111 (26.2)		1.34 (1.06–1.69)	0.013
Moderate (125–129 mEq/L)	230 (6.1)	38 (13.8)		2.54 (1.75–3.67)	<0.001	209 (5.8)	59 (14.0)		2.52 (1.88–3.39)	<0.001
Severe (<125 mEq/L)	141 (3.8)	21 (7.6)		3.00 (1.88–4.79)	<0.001	134 (3.7)	28 (6.6)		2.47 (1.65–3.69)	<0.001
p value for trend				<0.001					<0.001
Persistent hyponatremia			<0.001
Normal	2,560 (68.2)	138 (50.2)		1 (reference)		2,473 (68.5)	225 (53.2)		1 (reference)
Non-persistent	797 (21.2)	77 (28.0)		1.57 (1.18–2.09)	0.002	767 (21.3)	107 (25.3)		1.39 (1.09–1.76)	0.007
Persistent	399 (10.6)	60 (21.8)		2.29 (1.68–3.12)	<0.001	368 (10.2)	91 (21.5)		2.20 (1.72–2.83)	<0.001
p value for trend‡				<0.001					<0.001

HR, hazard ratio; CI, confidence interval.

^‡Multivariable model.

Results are expressed as number (column percentage) for bivariate comparisons and as multivariable-adjusted odds ratio and (95% confidence interval) for multivariable comparisons. Between-group comparisons performed using chi-square (bivariate) and Cox regression (multivariable). Multivariable analyses were performed adjusting for age (four categories), gender, main cause for hospitalization (eight categories), and number of associated comorbidities.