Skip to main content
. 2024 Jun 22;26(2):135–152. doi: 10.1016/j.ccrj.2024.03.002

Table 3.

Mixed effects hierarchical multivariable logistic regression for in-hospital mortality adjusted for sex, illness severity, COVID-19 status, frailty, ICU activity index and hospital type (with site as random effect) in all patients and in subgroups categorised by a. invasive therapies (invasive ventilation, renal replacement or ECMO), b. no invasive therapies and c. patients in public hospital ICUs.

Patient category CCRN group No. of patients Observed mortality Adjusted Odds Ratio (95% CI) p value
All patients CCRN >75% (n = 6563) 534 (8.1%) Reference value
CCRN 50–75% (n = 7695) 859 (11.2%) 1.21 (1.02–1.45) 0.032
CCRN <50% (n = 2360) 252 (10.7%) 1.21 (0.94–1.55) 0.14
Subgroups
a. Invasive ventilation, renal replacement or ECMO CCRN >75% (n = 2751) 352 (12.8%) Reference value
CCRN 50–75% (n = 3297) 602 (18.3%) 1.35 (1.11–1.64) 0.003
CCRN <50% (n = 952) 169 (17.8%) 1.28 (0.98–1.66) 0.07
b. No invasive ventilation, renal replacement or ECMO CCRN >75% (n = 3812) 182 (4.8%) Reference value
CCRN 50–75% (n = 4398) 257 (5.8%) 1.05 (0.79–1.39) 0.76
CCRN <50% (n = 1408) 83 (5.9%) 1.14 (0.76–1.71) 0.52
c. Public hospital ICUs CCRN >75% (n = 4177) 425 (10.2%) Reference value
CCRN 50–75% (n = 7274) 844 (11.6%) 1.27 (1.06–1.52) 0.011
CCRN <50% (n = 2350) 251 (10.7%) 1.24 (0.97–1.59) 0.08

For full multivariable models see Appendix Table 5. CCRN, Critical Care Registered Nurse; CI, Confidence interval; ECMO, Extracorporeal membrane oxygenation; ICU, Intensive Care Unit.