Frequency, associated factors and outcome of multi-drug-resistant intensive care unit-acquired pneumonia among patients colonized with extended-spectrum β-lactamase-producing Enterobacteriaceae

Annals of Intensive Care

Table 3 Cox regression (bivariable and multivariable) analyses of variables associated with death at sixty days

Variables	Bivariable analysis		Multivariable analysis
Variables	HR (95% CI)	p value	aHR (95% CI)	p value
SAPS2 > 43	1.76 (1.03–3.00)	0.038	1.93 (1.12–3.34)	0.018
Chronic pulmonary disease	1.68 (0.93–3.04)	0.086	–
Liver cirrhosis	1.89 (0.86–4.17)	0.11	–
Ab < 3 mo., broad-sp. > 10 d	2.21 (1.31–3.71)	0.003	–
C3G < 3 mo	1.64 (0.93–2.90)	0.087	–
Carbapenem < 3 mo	2.59 (1.11–6.06)	0.03	–
Charlson > 2	1.75 (1.04–2.95)	0.034	–
ESBL colonization at admission	1.56 (0.92–2.63)	0.10	–
Septic shock associated with nosocomial pneumonia	2.86 (1.68–4.85)	0.0001	2.81 (1.66–4.78)	<0.0001
VAP	0.48 (0.24–0.96)	0.037	0.48 (0.24–0.98)	0.04
ESBL-PE ICUAP	1.57 (0.93–2.64)	0.091	1.15 (0.65–2.05)	0.64
ICU-acquired infection before ICUAP	0.51 (0.28–0.95)	0.033	0.52 (0.28–0.97)	0.04
Others antibiotics between colonization and pneumonia	1.49 (0.89–2.52)	0.13	–
Appropriate empirical antimicrobial therapy^a	1.05 (0.56–1.95)	0.88	0.66 (0.34–1.27)	0.22

Ab antibiotic, broad-sp. broad-spectrum, 3GC third-generation cephalosporin; iBL beta-lactamase inhibitor, mo month, VAP ventilator-associated pneumonia, ICUAP ICU-acquired pneumonia, <3 mo within 3 months before ICU admission, HR (95% CI) hazard ratio interquartile range (25–75%)
^aAntibiotic treatment was considered adequate if one or more antibiotics initiated for ICUAP were active against the causative microorganism on the basis of the antibiotic susceptibility profile of the strain