LAUSR

The study by Sj€ oblom et al. [1] aimed to compare high-flow nasal oxygen (HFNO) with use of a tight facemask for preoxygenation during rapid sequence induction (RSI). While this question has undeniable merit, we are concerned that methodological issues, flawed outcomes and overlooking of real-world considerations may unjustifiably lead clinicians to conclude that HFNO is equivalent to facemask for pre-oxygenation, potentially placing patients at increased risk of harm. The study compares preoxygenation followed by apnoeic oxygenation (HFNO group) with pre-oxygenation alone (facemask group), preventing evaluation of the relative efficacy of each technique for pre-oxygenation. Pre-oxygenation and apnoeic oxygenation are distinct techniques for extending the safe apnoea time, and apnoeic oxygenation during laryngoscopy can be implemented independently of the pre-oxygenation technique. It is feasible that even if preoxygenation with HFNO was less effective than with facemask, the safe apnoea time in HFNO patients might still be longer, given subsequent apnoeic oxygenation in this group. Comparison of the efficacy of pre-oxygenation alone would require HFNO to be turned off during laryngoscopy so that apnoeic oxygenation was not performed in either group. The choice of oxygen desaturation < 93% as the primary outcome casts further doubt on the study’s conclusions because it is an insensitive measure of the benefits of pre-oxygenation. Patients may maintain oxygen saturations (SpO2) across a wide range of alveolar oxygen concentrations. It is possible that two patients who did not desaturate before tracheal intubation might still have vastly different oxygen reserves in their functional residual capacity. An ongoing interruption to alveolar oxygen delivery might reveal the residual safe apnoea time and SpO2 nadir (reflecting the effectiveness of pre-oxygenation) of these patients to be substantially different, despite them being deemed ‘equivalent’ with respect to the primary outcome of this study. Arguably the main benefit of preoxygenation is protecting against the severity and duration of exposure to hypoxaemia in such circumstances. The exclusion of patients most at risk of desaturation (BMI > 35 kg.m, pregnancy) and the inclusion of patients whose lungs were ventilated before tracheal intubation despite the absence of desaturation, limits the ability of this study to identify this benefit. End-tidal oxygen (ETO2) is not a “surrogate marker of the efficacy of pre-oxygenation” as the authors describe. Rather, it is a direct measure of denitrogenation of the functional residual capacity that defines pre-oxygenation. While factors such as alveolar dead space might influence the attained ETO2 and the availability of this oxygen reserve to the pulmonary capillaries in individual patients, such issues should be accounted for by adequate randomisation. Contrary to the authors’ assertion, it is possible to measure ETO2 following pre-oxygenation with HFNO. Although the technique for this would not have been practical in this study, Hanouz et al. [2] demonstrated in healthy volunteers that HFNO is not a reliable method for pre-oxygenation, even using a closedmouth technique. Consistent with this, the authors identify statistically and clinically significant higher ETO2 levels in the first breath after tracheal intubation in the facemask vs. HFNO groups. They suggest this could be explained by open-mouth breathing during pre-oxygenation, leading to a lower inspired oxygen concentration and lower residual lung oxygen stores in the HFNO group at tracheal intubation, despite apnoeic oxygenation. Such an explanation acknowledges that pre-oxygenation using HFNO is less effective and that the study outcomes are simply unable to expose the consequences of this. Their alternative explanation, that in theHFNOgroup the unused anaesthetic circuit may have contained only 21% oxygen, exposes a methodological flaw that prevents useful comparison of these values. It also reveals an important real-world consideration: unless denitrogenation of the anaesthetic circuit is undertaken, in the event of difficult tracheal intubation and desaturation, rescue ventilation in patients pre-oxygenated with HFNO will initially be with room air, delaying restoration of SaO2. Finally, this comparative study is underpowered, having failed to meet its predefined event rates. Therefore, the absence of a difference between groups does not support the authors’ conclusion that their study “proved that HFNO Anaesthesia 2021, 76, 1274–1282 Correspondence