Table 2.

Independent risk factors associated with candidemia according to hospitalization inside and outside intensive care units

Risk factors	Whole population1, 2 (N = 567)			Intensive care1, 2 (N = 250)			Non-Intensive care1, 2 (N = 322)
	OR	95% CI	p	OR	95% CI	p	OR	95% CI	p
Central venous catheter4	6.74	2.96–15.4	< 0.001				9.77	3.72–25.7	< 0.001
Total parenteral nutrition4	3.92	2.28–6.73	< 0.001	6.75	2.89–15.7	< 0.001	3.29	1.52–7.13	0.003
Previous septic shock	2.29	1.33–3.96	0.003	2.39	1.14–5.01	0.02
Acute kidney injury				4.77	1.94–11.8	< 0.001
Heart disease	1.78	0.96–3.33	0.07	3.78	1.09–13.1	0.006
Renal replacement therapy	2.16	1.11–4.21	0.02
Glycopeptides5, 6							3.31	1.33–8.23	0.01
Nitroimidazoles5, 6	2.16	1.05–4.45	0.04				3.12	1.07–9.11	0.04
Aminoglycosides5, 6				2.28	1.01–5.13	0.05

OR stands for odds ratio, CI for confidence interval

¹Variables in multivariate models were selected by stepwise regression, using a cutoff p value of 0.1. The number of patients in the model may be lower than the total number of patients due to missing co-variables in some individuals

²The models are not changed and the association with antibiotics is still significant when the variable “intraabdominal bacterial infection” is forced into the model

³SAPS2 was not included in the model since it is composed of variables which are presented separately in the model

⁴Within 2 weeks before candidemia (cases) or matched time period (controls)

⁵Within 4 weeks before candidemia (cases) or matched time period (controls)

⁶The association between these classes of antibiotics and candidemia is still significant when the variable “number of antibiotics” is added in the model (independent variables)