Influence of the trajectory of the urine output for 24 h on the occurrence of AKI in patients with sepsis in intensive care unit

Table 5 Results of double robust analysis

Models	Class1	Class2		Class3		Class4		Class5
Models		HR (95%CI)	P value	HR (95%CI)	P value	HR (95%CI)	P value	HR (95%CI)	P value
IPTW-Logistic
Propensity score IPTW	Reference	0.959 (0.509–1.807)	0.897	1.611 (1.206–2.151)	0.001	1.510 (1.101–2.072)	0.011	2.364 (1.830–3.054)	< 0.001
Doubly robust with unbalanced covariates	Reference	0.950 (0.515–1.752)	0.870	1.598 (1.195–2.137)	0.002	1.514 (1.116–2.054)	0.008	2.342 (1.811–3.027)	< 0.001
Doubly robust with all covariates	Reference	0.951 (0.532–1.699)	0.865	1.554 (1.163–2.075)	0.003	1.502 (1.120–2.013)	0.007	2.388 (1.853–3.079)	< 0.001
IPTW-XGboost
Propensity score IPTW	Reference	0.759 (0.411–1.402)	0.379	1.382 (1.034–1.849)	0.029	1.425 (1.063–1.912)	0.018	2.084 (1.613–2.693)	< 0.001
Doubly robust with unbalanced covariates	Reference	0.764 (0.412–1.418)	0.394	1.384 (1.035–1.851)	0.029	1.428 (1.064–1.917)	0.018	2.084 (1.612–2.694)	< 0.001
Doubly robust with all covariates	Reference	0.754 (0.424–1.343)	0.338	1.417 (1.059–1.896)	0.019	1.355 (1.011–1.817)	0.042	2.172 (1.676–2.816)	< 0.001

HR: hazard ratio; CI: confidence interval; IPTW: inverse probability of treatment weighting

ISSN: 1479-5876