PARIS – A massive randomized trial that included all New Zealanders with a suspected acute coronary syndrome during a 2-year period has provided definitive evidence that giving high-flow supplemental oxygen to those who are nonhypoxemic is of no clinical benefit, although it wasn’t harmful, either.
“Patients who have a normal blood oxygen saturation level are very unlikely to benefit from supplemental oxygen,”, said in presenting the results of the (New Zealand Oxygen Therapy in Acute Coronary Syndromes) trial at the annual congress of the European Society of Cardiology.
“It’s amazing that oxygen has been used in patients with suspected heart attack for over 50 years, and during that time there’s never been definite evidence that it improves outcomes. And more recently some have even suggested giving high-level oxygen might actually cause harm,” observed Dr. Stewart, a cardiologist at Auckland City Hospital and the University of Auckland (New Zealand).
The primary outcome in NZOTACS was 30-day all-cause mortality. In the overall study population, the rate was 3.0% in the group assigned to the routine high-flow oxygen protocol and closely similar at 3.1% in those randomized to the conservative oxygen strategy. And there was reassuringly no signal that the liberal oxygen protocol caused any harm.
To conduct this cluster randomized crossover trial, Dr. Stewart and his coinvestigators divided New Zealand into quadrants and, taking advantage of the coordinated health care systems operative in the nation of 4.8 million, they arranged for all ambulances, emergency departments, and hospitals in each geographic region to utilize each supplemental oxygen strategy for a total of 12 months.
In the liberal oxygen strategy, patients with suspected ACS on the basis of ischemic chest pain or ECG changes received high-flow oxygen by face mask at 6-8 L/min regardless of their blood oxygen saturation (SaO2) level. The oxygen was stopped only upon clinical resolution of myocardial ischemia. In contrast, in the low-oxygen protocol, supplemental oxygen was reserved for patients with an initial SaO2 below 90%, with a target SaO2 of 90%-94%.
Roughly 90% of the nearly 41,000 study participants had a normal SaO2 of 90% or more. Their 30-day mortality was 2.1% with the high-oxygen protocol and similar at 1.9% with the conservative oxygen protocol.
In contrast, there was a suggestion of benefit for the routine liberal oxygen strategy in the subgroup of patients with ST-elevation MI. Their 30-day mortality was 8.8% with high-flow oxygen and 10.6% with the conservative oxygen protocol. The resultant 19% relative risk reduction barely missed statistical significance. There was also a trend for possible benefit of routine high-flow oxygen in the roughly 12% of NZOTACS participants with an SaO2 below 95%, a lower bar than the 90% SaO2 that defines hypoxemia. Their death rate at 30 days was 10.1% if they got supplemental oxygen and 11.1% if they only received oxygen in the event their SaO2 was below 90%. But these exploratory findings must be viewed as hypothesis-generating, and a large confirmatory study would be required, Dr. Stewart noted.
Discussant, commented that, based on the NZOTACS results, he believes a couple of changes to the current ESC guidelines on management of ACS are in order. The guidelines now state that oxygen is indicated in patients with suspected ACS and hypoxemia as defined by an SaO2 below 90%, giving that recommendation a Class I Level of Evidence C. That should now be upgraded to the strongest-possible Class I A recommendation, according to Dr. Hofmann, a cardiologist at the Karolinska Institute in Stockholm.
The ESC guidelines also state that oxygen isn’t routinely recommended in patients with an SaO2 of 90% or more, rating that guidance Class III B. On the basis of NZOTACS coupled with earlier far smaller studies, that should be changed to a Class III A recommendation, meaning simply don’t do it. The hint provided by NZOTACS of a possible small benefit for oxygen in patients with an SaO2 below 95% isn’t strong enough evidence to carry the day, in Dr. Hofmann’s view.
Dr. Stewart and Dr. Hofmann reported having no financial conflicts of interest. The NZOTACS trial was funded by the National Heart Foundation of New Zealand.
SOURCE: Stewart R. ESC 2019, Hotline Session 2.