Skip to main content
. 2024 Mar 18;26:e50369. doi: 10.2196/50369

Table 3.

Baseline characteristics of the patients with sepsis following ICUa admission.

Variables MIMIC-IVb (n=20,933) eICU-CRDc (n=10,362)

Survivors (n=17,476) Nonsurvivors (n=3457) P value Survivors (n=8605) Nonsurvivors (n=1757) P value
Age (years), median (IQR) 67 (56-77) 71 (60-81) <.001 66 (54-76) 70 (60-80) <.001
Male, n (%) 10,285 (58.9) 1938 (56.1) .002 4606 (53.5) 952 (54.2) .62
BMI (kg/m2), median (IQR) 27.2 (23.8-31.7) 26.5 (23.1-31.3) <.001 27.3 (23.1-33.2) 25.9 (22.0-31.1) <.001
Ethnicity, n (%)

<.001

.97

Asian 508 (2.9) 100 (2.9)
140 (1.6) 27 (1.5)

Black 1739 (10) 315 (9.1)
868 (10.1) 181 (10.3)

White 11.972 (68.5) 2134 (61.7)
6699 (77.9) 1361 (77.5)

Other 3257 (18.6) 908 (26.3)
898 (10.4) 188 (10.7)
Comorbidities, n (%)

Chronic kidney disease 2913 (16.7) 673 (19.5) <.001 1040 (12.1) 277 (15.8) <.001

Congestive heart failure 5575 (31.9) 1332 (38.5) <.001 1907 (22.2) 429 (24.4) .04

Chronic pulmonary disease 4882 (27.9) 1055 (30.5) .002 2015 (23.4) 415 (23.6) .86

Liver disease 1209 (6.9) 487 (14.1) <.001 96 (1.1) 79 (4.5) <.001

Diabetes 5832 (33.4) 1103 (31.9) .09 2976 (34.6) 562 (32) .04
Pulmonary infection, n (%) 6958 (39.8) 1983 (57.4) <.001 4299 (50) 933 (53.1) .02
Vasopressors, n (%) 8836 (50.6) 2518 (72.8) <.001 2450 (28.5) 835 (47.5) <.001
Urine output (mL), median (IQR) 1640 (1030-2465) 956 (445-1715) <.001 1320 (689-2223) 700 (250-1351) <.001
Glasgow Coma Scale (score), median (IQR) 15 (13-15) 15 (12-15) .02 14 (9-15) 10 (6-14) <.001
APS IIId (score), median (IQR) 46 (35-60) 66 (51-84) <.001 51 (38-67) 71 (54-92) <.001
SOFAe (score), median (IQR) 5 (4-8) 8 (6-11) <.001 4 (2-7) 7 (4-10) <.001
Length of ICU stay (hours), median (IQR) 72 (44-138) 109 (58-208) <.001 68 (43-121) 86 (46-170) <.001

aICU: intensive care unit.

bMIMIC-IV: Medical Information Mart for Intensive Care database-IV.

ceICU-CRD: eICU Collaborative Research Database.

dAPS III: Acute Physiology Score III.

eSOFA: Sequential Organ Failure Assessment.