Association Between Early Hyperoxemia Exposure and Intensive Care Unit Mortality in Intracerebral Hemorrhage: An Observational Cohort Analysis

Supraphysiologic levels of oxygen could have potential adverse effects on the brain that may be dose and time dependent in patients with brain injury. We therefore aimed to assess whether exposure to excess supplemental oxygen, measured as time-weighted mean exposure to hyperoxemia, was associated with intensive care unit (ICU) mortality in patients with intracerebral hemorrhage (ICH).

In this single-center retrospective cohort study, we included all patients admitted to our ICU with a diagnosis of primary spontaneous ICH. To provide a longitudinal measure of hyperoxemia exposure, we calculated the hyperoxemia dose, defined as the area under the partial pressure of oxygen in arterial blood (PaO₂) time curve above the threshold PaO₂ value of 100 mm Hg (13.3 kPa) divided by the number of hours of potential exposure. To provide consistent potential exposure windows and limit bias from informative censoring, nested subsets were created with progressively longer exposure periods (0–1 day, 0–2 days, 0–3 days, 0–4 days, 0–5 days, 0–6 days, 0–7 days). We used multivariable Cox regression, with hyperoxemia dose as a time-dependent covariate, to model ICU mortality. Admission ICH and Acute Physiology and Chronic Health Evaluation II scores were included as predictor covariables. A step-function extended Cox model was also fitted.

Between September 2019 and July 2022, 275 patients met the inclusion criteria, with 24,588 arterial blood gas results available for analysis. The mean age was 57.19 years (± 13.99), 59.64% were male, 23.64% had an infratentorial origin of hemorrhage, and ICU mortality was 35.64%. Almost all patients (97.45%) were exposed to hyperoxemia during their ICU admission. Cox regression modeling showed an association between hyperoxemia dose and ICU mortality (hazard ratio 1.15, confidence interval 1.05–1.25, p = 0.003). This association was observed in the 0–1 day subset in the step-function extended Cox model (hazard ratio 1.19, confidence interval 1.06–1.35, p = 0.005) but not in any of the subsequent exposure periods.

In patients with ICH admitted to the ICU, we observed an association between hyperoxemia dose and ICU mortality. Further prospective study is required to inform guidance on early systemic oxygen targets in ICH.