Table 3.

Univariable and multivariable Cox proportional regression models explore the independent association of subphenotypes with 90-day mortality by including clinically meaningful covariates

	HR (95% CI)	p value	AIC	BIC
Univariable
Sex (Ref. M)	0.57 (0.41–0.81)	0.001	2175	2180
Any comorbidities (Ref. None)	3.26 (2.15–4.94)	< 0.001	2146	2151
Limitation of life sustaining measures (Ref. No)	28.73 (19.89–41.49)	< 0.001	1812	1816
Subphenotype 1 (Ref. Subphenotype 2)	3.49 (2.60–4.69)	< 0.001	2121	2125
Multivariable
Subphenotype 1 (adjusted for sex) (Ref. Subphenotype 2)	3.57 (2.65–4.79)	< 0.001	2110	2118
Subphenotype 1 (adjusted for any comorbidities) (Ref. Subphenotype 2)	3.07 (2.28–4.13)	< 0.001	2096	2105
Subphenotype 1 (adjusted for limitation of life sustaining measures) (Ref. Subphenotype 2)	1.60 (1.17–2.18)	0.003	1805	1814
Subphenotype 1 (adjusted for sex and any comorbidities) (Ref. Subphenotype 2)	3.12 (2.32–4.21)	< 0.001	2086	2099
Subphenotype 1 (adjusted for sex and limitation of life sustaining measures) (Ref. Subphenotype 2)	1.68 (1.23–2.30)	0.001	1803	1816
Subphenotype 1 (adjusted for any comorbidities and limitation of life sustaining measures) (Ref. Subphenotype 2)	1.54 (1.13–2.10)	0.006	1800	1813
Subphenotype 1 (adjusted for sex, any comorbidities and limitation of life sustaining measures) (Ref. Subphenotype 2)	1.63 (1.19–2.23)	0.002	1797	1815

Mortality risk was reported by hazard ratio with 95% confidence interval. Adjusted models were ranked by their Akaike information criterion (AIC) and their Bayesian information criterion (BIC). N = 549