Time-varying intensity of oxygen exposure is associated with mortality in critically ill patients with mechanical ventilation

Table 2 Effect of time-varying hyperoxemia and high FiO₂ on 28-day mortality of 7784 patients with mechanical ventilation using PAMMs

	Exposure to hyperoxemia		Exposure to high FiO₂
	HR (95% CI)	P value	HR (95% CI)	P value
Baseline variables
Age, years	1.018(1.014–1.022)	< 0.001	1.018(1.014–1.022)	< 0.001
Male	0.981 (0.892–1.078)	0.69	0.972 (0.885–1.069)	0.57
Admission type
Medical	Reference		Reference
Surgical elective	0.248 (0.158–0.389	< 0.001	0.246 (0.157–0.386)	< 0.001
Surgical urgent	0.846 (0.759–0.943)	0.024	0.834 (0.748–0.929)	< 0.001
Other	0.643 (0.558–0.742)	< 0.001	0.642 (0.557–0.741)	< 0.001
Weight, kg	0.993 (0.990–0.994)	< 0.001	0.992 (0.990–0.994)	< 0.001
Elixhauser comorbidity score	1.033 (1.014–1.052)	< 0.001	1.038 (1.019–1.057)	< 0.001
Time-varying variables
Receiving IMV	0.417 (0.375–0.463)	< 0.001	0.437 (0.393–0.486)	< 0.001
Use of vasopressor	1.370 (1.223–1.534)	< 0.001	1.380 (1.232–1.546)	< 0.001
SOFA score	1.145 (1.129–1.160)	< 0.001	1.131 (1.116–1.147)	< 0.001
PaCO₂, mmHg	1.016 (1.013–1.020)	< 0.001	1.014 (1.010–1.017)	< 0.001
Any hyperoxemia (TWA-PaO₂ ≥ 120 mmHg)	1.166 (1.059–1.284)	0.0017	–	–
Any exposure to high FiO₂ (TWA-FiO₂ ≥ 0.5)	–	–	1.496 (1.363–1.641	< 0.001

PAMMs Piece-wise exponential additive mixed models; HR Hazard ratio; CI Confidence interval; IMV Invasive mechanical ventilation; SOFA sequential organ failure assessment; TWA Time-weighted average; PCO₂ partial pressure of carbon dioxide; FiO₂ Fraction of inspired oxygen

ISSN: 1364-8535