Oxygenation metrics have limited prognostic value in non-asphyxial hypothermic cardiac arrest.

Abstract

Arterial blood gas analysis is a key component of the initial assessment for patients with cardiac arrest. In accidental hypothermia, temperature effects on gas solubility and hemoglobin-oxygen affinity confound oxygenation status. It remains unclear whether these values can be used to exclude patients from extracorporeal cardiopulmonary resuscitation in hypothermic cardiac arrest. To determine whether initial arterial oxygenation parameters predict favorable neurological outcomes in patients with hypothermic cardiac arrest treated with extracorporeal cardiopulmonary resuscitation. We conducted a retrospective analysis of data from a national registry of patients with severe accidental hypothermia treated between 2014 and 2025. The study included adult patients with core temperature ≤28 °C who underwent extracorporeal cardiopulmonary resuscitation for hypothermic cardiac arrest. Arterial blood samples drawn on arrival, before extracorporeal support, were temperature-corrected using established equations. The primary outcome was survival to hospital discharge with good neurological status (Cerebral Performance Category 1-2). Associations were evaluated with multivariable generalized additive models to allow non-linear effects. In our cohort of 89 patients, neither arterial oxygen partial pressure nor saturation were independent predictors of a favorable neurological outcome. When corrected for temperature, severe hypoxemia <50 mmHg was present in two-thirds of patients with a good neurological outcome. A single arterial blood gas analysis is unreliable for determining prognosis in patients with non-asphyxial hypothermic cardiac arrest. Initial hypoxemia does not preclude a favorable neurological outcome; therefore, oxygenation metrics should not serve as sole exclusion criterion for ECPR in hypothermic cardiac arrest.