Before the ICU: does emergency room hyperoxia affect outcome?

There is now ample evidence that hyperox(em)ia—that is, increased inspired oxygen concentrations (FIO2) and the subsequent rise in arterial oxygen tensions (PaO2)—coincides with aggravated mortality [1]. Most of the data originate from retrospective analyses, but a single-center trial showed that “conservative” PaO2 (70–100 mmHg) halved mortality when compared to “conventional” targets (≤ 150 mmHg) [2]. The available studies mostly refer to data from intensive care unit (ICU) patients, but despite its frequent use in daily practice, the impact of hyperox(em)ia remains much less clear for patients in the emergency department (ED) and/or even prior to hospital admission. Hyperox(em)ia is often present after initiation of mechanical ventilation, most likely for fear of hypoxemia when blood gas analyses are not readily available. However, supplemental O2 can also yield hyperoxemic PaO2 levels without mechanical ventilation: in the aforementioned clinical trial demonstrating the beneficial effect of targeting “conservative” PaO2 levels in the ICU, upon admission into the study only 2/3 of the patients investigated were mechanically ventilated [2]. However, the duration of mechanical ventilation per se is directly related to adverse outcome in ED patients. Mechanical ventilation in the ED is mostly initiated upon the necessity for airway management, in particular in the unconscious patient (e.g., in the context of intoxication, metabolic crises, and/or traumatic brain injury (TBI)), respiratory failure (e.g., pneumonia and/or exacerbation of chronic obstructive pulmonary disease (COPD)), circulatory shock, and/or after cardiac arrest. Hence, the question arises: depending on the underlying conditions, does hyperox(em)ia affect the outcome of patients in the ED, in particular when they require mechanical ventilation? It is well established that hyperoxemia (defined as a PaO2 > 100 mmHg) is associated with adverse outcome in patients necessitating mechanical ventilation due to exacerbation of chronic lung disease (i.e., asthma or COPD) [3]. While there are no clinical studies on the impact of hyperox(em)ia in patients with community-acquired pneumonia, a recent retrospective study in this journal showed that hyperoxemia (defined as PaO2 > 120 mmHg) increased the risk of ventilator-associated pneumonia in patients receiving mechanical ventilation for more than 48 h [4]. The recent HYPER2S trial yielded deleterious effects of hyperoxemia in patients with septic shock (44% of pulmonary origin): FIO2 = 1.0 during the first 24 h after initial hemodynamic stabilization increased mortality at days 28 and 90 despite a significantly lower sequential organ failure assessment (SOFA) index at day 7, but without affecting the rate of secondary pneumonia or infection in general [5]. During the acute phase of circulatory shock, “the administration of oxygen should be started immediately to increase oxygen delivery and prevent pulmonary hypertension” [6]. The results of the HYPER2S trial suggest that hyperox(em)ia is deleterious in situations of distributive shock where “the main deficit lies in the periphery, with ... altered oxygen extraction” [6]. What about shock characterized by low cardiac output and, hence, inadequate oxygen transport? While there are no data on the outcome effects of hyperox(em)ia in cardiogenic shock, it is well established that it increases systemic vascular resistance in patients with congestive heart failure [7]. In line with this, two large randomized, controlled trials have shown that hyperoxemia started already during the prehospital phase offers no survival benefit at all [8] and can even increase mortality [9] in patients with acute myocardial infarction, possibly to the preferential vasoconstrictor effect of oxygen in the coronary circulation [7]. In contrast, the role of hyperox(em)ia during hypovolemia, in particular due to trauma and hemorrhage, is much less clear: due to the blood loss-related drop in oxygen transport capacity, hyperox(em)ia is frequently used to restore tissue oxygen * Correspondence: [email protected] Institut für Anästhesiologische Pathophysiologie und Verfahrensentwicklung, Universitätsklinikum Ulm, Helmholtzstrasse 8/1, 89081 Ulm, Germany Full list of author information is available at the end of the article