Determining delayed admission to the intensive care unit for mechanically ventilated patients in the emergency department

Table 2 Effects of ICU waiting time on mortality in the first 21 ventilator days by subsets of different waiting hours

Population subset by different ICU-waiting hours	HR^a(95% CI)	Pvalue
Less than 2 hours (n = 118)	0.64 (0.28-1.48)	0.297
Less than 3 hours (n = 240)	1.25 (0.92-1.71)	0.147
Less than 4 hours (n = 337)	1.17 (0.98-1.39)	0.093
Less than 5 hours (n = 420)	1.15 (1.02-1.31)	0.026
Less than 6 hours (n = 473)	1.14 (1.03-1.26)	0.008
Less than 7 hours (n = 524)	1.13 (1.04-1.21)	0.002
Less than 8 hours (n = 561)	1.11 (1.04-1.18)	0.001
Less than 9 hours (n = 591)	1.07 (1.01-1.13)	0.017
Less than 10 hours (n = 622)	1.03 (1.00-1.05)^b	0.043
Less than 11 hours (n = 657)	1.02 (0.98-1.06)	0.354
Less than 12 hours (n = 681)	1.01 (0.97-1.04)	0.772
Less than 24 hours (n = 879)	0.99 (0.98-1.01)	0.436
Entire waiting hours (n = 1242)	0.99 (0.99-1.00)^b	<0.001

^aHR of ICU waiting hour, model of multivariate Cox regression, adjusted for sex, ICU waiting hour, APACHE II score, triage, unplanned ED revisit in 72 hours and readmission in 14 days; ^bnot satisfied by the proportional assumption of Cox model, assessed by Kolmogorov-type supremum test. ICU, intensive care unit; HR, hazard ratio; CI, confidence interval; ED, emergency department; APACHE, acute physiology and chronic health evaluation.

ISSN: 1364-8535