Adiponectin, retinol-binding protein 4, and leptin in protracted critical illness of pulmonary origin

Table 2 Baseline and outcome characteristics of septic versus nonseptic critically ill patients (ICU stay ≥5 days)

	Proven sepsis (n = 165)	Proven no sepsis (n = 94)	P
Sex (number [percentage] male])	111 [67.3]	59 [62.8]	0.5
Age (years; mean ± SD)	66.1 ± 12.3	67.0 ± 15.2	0.6
BMI (mean ± SD)	25.1 ± 5.1	24.1 ± 5.0	0.1
Admission Apache II score (median [IQR])	23 [17–29]	21 [17–28]	0.6
Admission SOFA score (median [IQR])	7 [4–9]	4 [3–7]	< 0.0001
Kidney failure on admission (number [percentage])	27 [16.3]	12 [12.8]	0.4
History of diabetes (number [percentage])	20 [12.1]	15 [16.0]	0.4
Blood glucose on admission (mg/dl; mean ± SD)	164.5 ± 66.6	166.7 ± 70.3	0.8
Maximum SOFA score during ICU stay (mean ± SD)	9.4 ± 4.0	7.1 ± 3.7	< 0.0001
Days in ICU (median [IQR])	12 [7–20]	12.5 [7–19]	0.5
Death in ICU (number [percentage])	54 [32.7]	24 [25.5]	0.2
In-hospital deaths (number [percentage])	85 [51.5]	41 [43.6]	0.2

All subjects were screened for infection on admission and were defined retrospectively for this post hoc study as "proven sepsis" when infection was proven and the Bone criteria were fulfilled [29]. Patients proven not to have sepsis with this strict definition were classified as "proven no sepsis". Fifty-nine patients for whom treatment with antibiotics was started before admission to the ICU were categorized as unknown and therefore excluded from this subanalysis.

ISSN: 1364-8535