乌鸦传媒

Treating acute respiratory distress syndrome (ARDS) with conservative oxygen targets was not helpful and possibly harmful, the LOCO₂ trial showed.

Mortality at day 28 was a nonsignificant 7.8 percentage points higher (34.3% vs 26.5%) with conservative targets of a 55-70 mm Hg partial pressure of arterial oxygen (PaO₂) and oxygen saturation by pulse oximetry (SpO₂) of 88% to 92% compared with a liberal strategy targeting 90-105 mm Hg PaO₂ and ≥96% SpO₂.

A secondary endpoint, 90-day mortality, was significantly worse with the conservative targets (44.4% vs 30.4%), reported Gilles Capellier, MD, PhD, at Centre Hospitalier Régional Universitaire Besançon, France, and colleagues in the.

Aside from that "worrisome" signal, five mesenteric ischemic events occurred in the conservative group compared with none in the liberal-oxygen group. The trial was stopped early after enrollment of 205 patients due to futility and the safety concerns.

"Although decreasing oxygen exposure (lower FiO₂ [fraction of inspired oxygen]) might decrease the lung damage at the early phase of the disease, patients were exposed to hypoxemia," the researchers wrote.

A second trial of conservative oxygen in a broader mechanically ventilated population, dubbed ICU-ROX, appeared in the same issue of the journal, although in October 2019. Its conservative oxygen saturation target of 90% to 96% did not lead to earlier weaning versus usual care and had a slight numeric excess of 180-day mortality.

Neither trial was large enough to rule out a null effect and more nuanced is needed, Derek Angus, MD, MPH, of the University of Pittsburgh School of Medicine, noted in an accompanying .

Meanwhile, "avoiding excess oxygen (i.e., not administering supplemental oxygen when the SpO₂ is 96% or greater and not starting supplemental oxygen when the SpO₂ is 92% or 93%) seems sensible, as per recent guidelines," he wrote.

Based on the potential risk seen in LOCO₂, "the lower range of the SpO₂ target in any conservative strategy, especially in patients requiring a high level of FiO₂, should perhaps be 90%, as was used in ICU-ROX, rather than 88%," he concluded.

Indeed, "combining these two studies and the , I am recommended in my unit now to target an SpO₂ range of 94% to 96%," Capellier told 乌鸦传媒.

A transiently lower oxygen saturation might have to be tolerated, such as when there's no extracorporeal membrane oxygenation support available or poor tolerance to high positive end-expiratory pressure, he added. "But as soon as possible, you might aim for a higher range."

There are implications too for COVID-19 coronavirus, because ARDS is present in more than 50% of these patients admitted to the ICU, Capellier noted. "Our recommendations clearly apply to these groups of patients."

The open-label trial was conducted in 13 ICUs in France over the first 7 days of invasive mechanical ventilation or until extubation if done before day 7.

Enrolled patients were intubated and had been on mechanical ventilation for less than 12 hours for ARDS, defined according to the Berlin definition, with less than 7 days between the lung damage and new or worsening respiratory symptoms. Slightly more than 40% of the participants had a PaO₂:FiO₂ lower than 100 mm Hg.

Long-term oxygen therapy or noninvasive ventilation at home, cardiac arrest, traumatic brain injury, and cranial hypertension were other exclusion criteria. Bilateral opacities on chest imaging and respiratory failure could not be fully accounted for by heart failure or fluid overload.

Comment