现在的位置: 首页研究点评, 进展交流>正文
[NEJM发表述评]:氧疗对ICU病死率的影响
2021年05月11日 研究点评, 进展交流 暂无评论

EDITORIAL

Effect of Oxygen Therapy on Mortality in the ICU

Paul J. Young

N Engl J Med 2021; 384:1361-1363
DOI: 10.1056/NEJMe2101538

Oxygen therapy is a key component of supportive care for patients who have hypoxemic respiratory failure and are being treated in the intensive care unit (ICU). Yet, although oxygen can be lifesaving for patients with severe hypoxemia, overzealous oxygen administration may be harmful. Furthermore, data to inform the use of oxygen therapy in patients with acute hypoxemia have been limited. In a 2018 systematic meta-analysis of randomized, controlled trials of oxygen use in acutely ill adults,1 approximately a third of all mortality data showing a benefit for conservative oxygen therapy were from a single-center study.2 The cited study, which was stopped at an unplanned interim analysis, showed an absolute difference in ICU mortality of 8.6 percentage points favoring conservative oxygen therapy. Only one other trial (a pilot evaluation) focusing on patients in the ICU was included in the meta-analysis.3

More recently, two randomized trials investigating oxygen regimens in the ICU were published in the Journal. The first trial, the Intensive Care Unit Randomized Trial Comparing Two Approaches to Oxygen Therapy (ICU-ROX), showed similar 180-day mortality among patients receiving invasive mechanical ventilation who were assigned to a conservative or liberal oxygen strategy.4 The second trial, the Liberal Oxygenation versus Conservative Oxygenation in Acute Respiratory Distress Syndrome (LOCO2) trial, was stopped early because of concern about intestinal ischemia events and deaths in the conservative-oxygen group.5

In this issue of the Journal, investigators report the results of the Handling Oxygenation Targets in the ICU (HOT-ICU) trial,6 which provides the next piece of the puzzle. In this trial, adults with acute hypoxemic respiratory failure were assigned to receive oxygen therapy targeting an arterial partial pressure of oxygen (Pao2) of 60 mm Hg (lower-oxygenation group) or 90 mm Hg (higher-oxygenation group). In the lower-oxygenation group, the median reported Pao2 was approximately 10 mm Hg higher than the target (70.8 mm Hg; interquartile range, 66.6 to 76.5), whereas the median value was closer to the target in the higher-oxygenation group (93.3 mm Hg; interquartile range, 87.1 to 98.7). The incidence of the primary outcome, mortality at 90 days, did not differ significantly between the two groups (42.9% in the lower-oxygenation group and 42.4% in the higher-oxygenation group, P=0.64). The percentage of days that patients were alive without life support, days alive after hospital discharge, and adverse events, including intestinal ischemia, were also similar.

Determining what these data mean for clinical practice is complex. One consideration is that the Covid-19 pandemic has resulted in an unprecedented demand for oxygen therapy. In ICUs where oxygen supplies are constrained, the data from the HOT-ICU trial provide a reasonable basis for a decision to implement conservative oxygen therapy. With respect to the safety of this approach, it is reassuring that the increased risk of intestinal ischemia with conservative oxygen therapy that was suggested in the LOCO2trial was not observed in the HOT-ICU trial. Since enrollment in the HOT-ICU trial was approximately 15 times that in the LOCO2 trial and the LOCO2 trial was stopped at an unplanned interim analysis, the most likely explanation for these discordant findings is that the unequal distribution of intestinal ischemia between groups in the LOCO2 trial occurred by chance.

In the absence of constraints in oxygen supply, on the basis of the HOT-ICU data, there is no particular reason to shift toward lower oxygenation targets in patients with hypoxemic respiratory failure. However, it is important to note that the HOT-ICU trial was designed to test the hypothesis that conservative oxygen therapy would reduce 90-day mortality by an absolute margin of 5 percentage points. Although an effect of this magnitude (either beneficial or harmful) now appears to be very unlikely, the absence of evidence for such a large mortality effect should not be misconstrued as evidence of the absence of a smaller, but still clinically important, effect on mortality. The 95% confidence interval around the 90-day effect on mortality encompasses the possibility that conservative oxygen therapy could result in either an absolute reduction of 2.9 percentage points in mortality or an absolute increase of 4.2 percentage points. Although there is no established minimal clinically important difference in 90-day mortality for ICU patients, reducing mortality by even 1.5 percentage points would save 1500 lives for every 100,000 patients treated. Conducting a trial that is large enough to detect such a small effect will be challenging,7 but given how widely oxygen is used, there is a public health imperative to do so.

The risks and benefits of oxygenation targets may differ according to the causes of hypoxemic respiratory failure and patients’ coexisting illnesses. Some groups of patients, including surgical patients and those with acute brain injuries, were not well represented in the HOT-ICU trial. Even in prespecified subgroups, the 95% confidence intervals around the mortality outcomes were wide and were consistent with a range of plausible treatment effects.8 Future trials that include a broad range of patients who receive oxygen therapy in the ICU and that are designed to detect between-group differences in 90-day mortality on the order of 1.5 percentage points should provide considerable power to detect a heterogeneity of treatment effects in important subgroups.7 Until such trials are completed, the HOT-ICU trial data support a position of equipoise about conservative versus liberal oxygen therapy in patients with hypoxemic respiratory failure in the ICU.

给我留言

您必须 [ 登录 ] 才能发表留言!

×
腾讯微博