Identification of persistent and resolving subphenotypes of acute hypoxemic respiratory failure in two independent cohorts

Table 3 Relative risk of mortality associated with persistent and resolving hypoxemic respiratory failure, stratified by acute respiratory distress syndrome (± ARDS)

		Total	Deaths	Relative risk (95% confidence interval)
		N	N (%)	Unadjusted	Model A	Model B
Discovery cohort
Resolving	−ARDS	296	22 (7%)	1.00 (reference)	1.00 (reference)	1.00 (reference)
Resolving	+ARDS	86	9 (10%)	1.41 (0.67, 2.94)	1.36 (0.65, 2.83)	0.96 (0.49, 1.89)
Persistent	−ARDS	186	28 (15%)	2.03 (1.19, 3.43)^b	2.03 (1.20, 3.43)^b	1.48 (0.86, 2.52)
Persistent	+ARDS	200	49 (25%)	3.30 (2.06, 5.28)^a	3.09 (1.91, 5.02)^a	1.81 (1.07, 3.08)^c
Validation cohort
Resolving	−ARDS	515	37 (7%)	1.00 (reference)	1.00 (reference)	1.00 (reference)
Resolving	+ARDS	168	33 (20%)	2.73 (1.77, 4.23)^a	2.48 (1.53, 4.02)^a	2.50 (1.65, 3.80)^a
Persistent	−ARDS	559	114 (20%)	2.84 (2.00, 4.03)^a	2.59 (1.77, 3.79)^a	2.52 (1.78, 3.58)^a
Persistent	+sARDS	473	116 (25%)	3.41 (2.41, 4.84)^a	3.19 (2.17, 4.69)^a	2.93 (2.07, 4.16)^a

Mortality is in-hospital mortality 28 days after enrollment
+ARDS refers to patients who were adjudicated as ARDS at any point by ICU day 3; −ARDS are patients who were not adjudicated to have ARDS by ICU day 3
Model A: adjusted for age, sex, chronic respiratory disease, and PaO₂-to-FIO₂ ratio on enrollment
Model B: adjusted for age, sex, chronic respiratory disease, and modified acute physiology and chronic health evaluation on enrollment (APACHE-III score in discovery cohort and APACHE-II score in validation cohort)
^ap < 0.001 ^bp < 0.01 ^cp < 0.05

ISSN: 1364-8535