NEJM Published online 8 April, 2021 | https://www.nejm.org/doi/full/10.1056/NEJMoa2032510
Digest author: Emer Kelly, incoming ERS E-Learning Director / 23 April, 2021
In recent years the question of the appropriate target range for oxygenation in the ICU has been investigated. Although a systematic review and meta-analysis showed that lower oxygenation targets were preferable in acutely ill adults, the LOCO2 trial in ARDS raised concern when stopped prematurely due to increase in mesenteric ischaemia when lower oxygen was targetted. The ICU-ROX trial found no between-group differences in ventilator-free days or in mortality within 28 days. To add to these major trails published last year, this study (HOT-ICU trial) tested the hypothesis that targeting a PaO2 of 60mmHg (8 kPa) would reduce 90 day mortality by 5% as compared to PaO2 of 90mmHG (12 kPa) in patients admitted to the ICU with hypoxemic respiratory failure.
This was an investigator-initiated, multicentre stratified, parallel-group clinical trial with centralized randomization and a computer-generated concealed assignment sequence, which stratified according to the trial site and the presence of COPD or active haematological malignancy. The nature of the study meant that this could not be carried out in a blinded fashion. Thirty-five ICUs in Europe were involved.
Patients >18 years old, admitted to ICU with hypoxemic respiratory failure, requiring at least 10 litres of oxygen per minute in an open system or FiO2 of at least 0.5 in a closed system were screened if they were expected to be in ICU for >24 hours. Patients were excluded if they could not be randomised within the first 12 hours in the ICU. Patients needed an arterial line and oxygenation targets were achieved by adjustment of FiO2. Saturations were used to guide the oxygenation once correlated with the ABG. The primary outcome was death within 90 days. Secondary outcomes included serious adverse events especially ischaemic events.
2928 were included; 1462 in the lower-oxygen group and 1466 in the higher oxygenation group. Importantly the recorded PAO2 measurements were lower in the lower oxygenation group than the higher. Ventilation approaches, circulatory supports, renal replacement and blood transfusions were similar between groups. At 90 days 618 (42.9%) in the lower-oxygenation group and 613 of 1447 patients (42.4%) in the higher oxygenation group had died (risk ratio 1.02). There was no difference in secondary outcomes: no difference in percentage of days alive without life support, percentage of days alive after hospital discharge or serious adverse events.
These findings provide some further evidence for the use of conservative oxygen therapy in patients with hypoxemic respiratory failure and does not raise the safety concerns raised by the LOCO2 trial. Overall mortality was twice as high in these ICU patients as had been hypothesised and was likely due to the nature of the admissions with a higher proportion of acute medical conditions. A systematic review up to the point of inclusion of the ICU-ROX trial showed neither beneficial nor harmful effects of higher versus lower oxygenation strategies. This study adds to the literature by finding no differences between the 2 oxygenation groups but still does not preclude the possibility of clinically important harm or benefit.