Apnoeic oxygenation is not novel. Three-hundred and fifty years ago, Robert Hook elegantly demonstrated that cyclical movements of the thoracic cavity and lungs were not necessary to keep a dog alive for prolonged periods. After removing the ribs and diaphragm and placing a catheter in the transected trachea, a constant flow of fresh air was supplied by a bellows “by which means the lungs also were always kept very full, and without any motion” 1. As long as the constant flow of air persisted, so did the dog's regular heartbeat. One hundred years ago, Nagel kept curarised pigeons alive by sending a stream of warmed air into the air sacs 2, whereas Meltzer and Auer observed dogs for over four hours with no respiratory movements 3. Nearly 300 years after Hook's first experiments, the idea of apnoeic oxygenation was translated into anaesthetised humans 4, and the last 50 years have seen regular investigations to understand the physiological principles of this technique, its relationship to ventilation and its clinical applications 5. Of note, apnoeic oxygenation refers to the delivery of oxygen to the alveoli with no obvious carbon dioxide clearance, whereas apnoeic ventilation demonstrates the delivery of oxygen to the alveoli with carbon dioxide clearance. In this issue of Anaesthesia, O'Loughlin et al. further build on this evidence base by reporting an observational study of 64 patients undergoing laryngeal surgery under general anaesthesia receiving apnoeic oxygenation by means of a tracheal catheter delivering 0.5–1.0 l.min−1 oxygen 6. This technique allowed 97% of cases to be successfully completed in just under 20 min, and 90% of cases proceeded without arterial oxygen desaturation. Interestingly, the rise in the end-tidal and venous partial pressure of carbon dioxide was similar 7 or marginally lower 8-10 than published data in modern practice as 0.12 and 0.15 kPa.min−1, respectively. This interesting and well-designed study not only adds to our knowledge and supports previous data, but it also highlights to practitioners that simple and age-old techniques remain valid. However, as the authors rightly state, their study suffers from some limitations. The drawbacks of observational studies are not lost on the authors: the fact that their patients were reasonably healthy with a mean body mass index of 25 kg.m−2 hamper the generalisability of these results to the type of patient we are all more likely to encounter; and they used peripheral venous blood samples only at the beginning and at the end of apnoea to determine their carbon dioxide trends. The study by O'Loughlin et al. also suffers other limitations that readers must also consider. Firstly, the study is not entirely novel. They aimed to confirm the results of Rudlof et al. in a prospective manner 9, but the use of tracheal catheters for apnoeic oxygenation is well-reported elsewhere. Secondly, this technique, by definition, risks barotrauma, depending on: the oxygen flow rate used; the size of the catheter inserted; the position of the catheter tip either centrally in the trachea or within the bronchi; and partial or total occlusion to the egress of gas. When using this technique, care must be taken to ensure the entire airway remains patent. Third, the use of this technique for laryngeal surgery also precludes the use of lasers. In the setting of laryngeal surgery, the utility of lasers are increasingly recognised, but the risks remain under-appreciated 11; thus, caution must always be exercised. And finally, this technique is reserved for intra-operative oxygenation, as opposed to other oxygenation techniques, such as high-flow nasal oxygen 7, 12, 13, that can be used for pre- and per-oxygenation. These latter two uses of apnoeic oxygenation (the ability to maximise oxygen storage in the lungs before apnoea and replenish oxygen within the alveoli during apnoea) are perhaps the most important, as they transcend the realms of tertiary centre head and neck surgical care and are applicable to all patients undergoing general anaesthesia. That being said, O'Loughlin et al. provide a timely opportunity for us to further explore the role of pre-oxygenation, flow rates and the anatomical location of oxygen delivery to maximise apnoeic oxygenation and ventilation. Pre-oxygenation is the process of delivering 100% oxygen to a spontaneously ventilating patient to denitrogenate the functional residual capacity of the lungs, providing an oxygen store. This maximises the amount of apnoeic time following induction of general anaesthesia before arterial oxygen desaturation occurs. However, both the efficacy as well as efficiency are important considerations, and they are often inappropriately used interchangeably 14, 15. Efficacy describes the process of pre-oxygenation, typically to the surrogate endpoint of an end-tidal oxygen concentration of > 90%. Efficacy is affected by: the anaesthesia circuit used; the oxygen flow rate applied; the oxygen concentration delivered through that circuit; an adequate facemask seal; presence of a leak anywhere within the circuit; the duration and pattern of breathing (tidal volume or deep breathing); and the alveolar ventilation:functional residual capacity ratio. However, it is efficiency that is ultimately the aim of pre-oxygenation; that is, to prolong the apnoea time after the induction of general anaesthesia (Fig. 1). Factors that affect efficiency (how long before arterial oxygen desaturation occurs) include: the volume of oxygen in the lungs at the end of pre-oxygenation; patient positioning affecting functional residual capacity; apnoeic oxygenation; apnoeic ventilation 7; and systemic oxygen consumption 14. The distinction between efficacy and efficiency during pre-oxygenation can be highlighted by considering the term parturient. Efficacy in pre-oxygenation could potentially be rapid and high due to high alveolar ventilation and reduced functional residual capacity, leading to rapid achievement of end-tidal oxygen concentrations > 90% with a facemask 16, 17. However, the efficiency of pre-oxygenation is poor due to reduced functional residual capacity and increased systemic oxygen consumption, leading to more rapid arterial oxygen desaturation 14. Attempts to improve the efficiency of pre-oxygenation and extend the apnoeic period have driven the research to better understand the optimal site for the delivery of oxygen. Some of these techniques, such as that reported by O'Loughlin et al., have allowed a significant extension of apnoea time with the possibility of laryngeal surgery, whilst others are techniques to increase the apnoea time in at risk groups during airway management, laryngoscopy and tracheal intubation. The location of oxygen delivery can be described anatomically from the most proximal position within the airway (nose) to the most distal (trachea or bronchi) and include: directly into the nose with simple low-flow nasal cannula 18; directly into the nose with high-flow nasal cannula 7, 10; into the nasopharynx with nasal catheters 19-21; into the oropharynx as buccal tube oxygen administration 22; and into the trachea or bronchi via a catheter 6, 9. Teller et al. were the first to describe insufflation of oxygen 3 l.min−1 into the pharynx of apnoeic and paralysed patients using an 8-Fr catheter located 2 cm distal to the tip of a nasal airway, and revealed a significant delay in arterial oxygen desaturation 21. In obese patients, oxygen delivered at 5 l.min−1 through a 10-Fr catheter inserted into the nasopharynx delayed the onset of apnoea during the 4-min study period 19. When difficult laryngoscopy was simulated, oxygen delivered at 5 l.min−1 through nasal cannulae was shown to maintain a higher arterial oxygen saturation during a 6-min study period 23. In a randomised controlled study, Achar et al. investigated the influence of oxygen delivery at 5 l.min−1 either proximally via nasal prongs or more distally via a nasopharyngeal catheter over a 10-min simulated prolonged difficult laryngoscopy scenario. They found that 9/28 (32%) patients with the proximal nasal cannula desaturated whilst none in the more distal nasopharyngeal catheter group desaturated 20. In terms of efficacy, pre-oxygenation is uncommonly achieved with oxygen delivery anywhere below the oropharynx, and thus is generally unaffected by the location of oxygen delivery. However, the combined findings of all of these studies shows that delivery of oxygen anywhere within the upper airway during the apnoeic period improves the safe apnoea time (time to arterial oxygen desaturation), thus efficiency, but the more distal the delivery the greater the influence on apnoea time. Practitioners should note that the more distal the delivery, the more care needs to be taken due to the greater risk for potential barotrauma. Put simply, the lower you go, the lower the flow. The effect of nasal high-flow oxygen has been modelled with computer simulation and using particle image velocimetry, viewed gas clearance and intratracheal turbulence measurements using three-dimensional models 24. This modelling and simulation evidence has demonstrated that flow rate is an important factor in prolonging safe apnoea time. Similarly, clinical data have shown a comparable extension of apnoea times during laryngeal surgery 7, 8, 10. Mullen et al. investigated the influence of flow rates of 0 l.min−1, 15 l.min−1 and 60 l.min−1 via nasal cannulae during 9 min of apnoea in patients with a body mass index of 28–35 kg.m−2. They reported that all of the patients receiving 0 l.min−1 had arterial oxygen desaturation to < 95% within the 9 min. However, the arterial partial pressure of oxygen was significantly lower in the 15 l.min−1 group compared with the 60 l l.min−1 group throughout the study period. Moreover, more patients receiving oxygen at 60 l.min−1 maintained adequate oxygenation for the entire 9-min duration 25. Similarly, Ricottilli et al. investigated the influence of high-flow nasal oxygen at 70 l.min−1 on maintenance of functional residual capacity and subsequent arterial oxygen desaturation following induction of general anaesthesia in obese patients compared with facemask oxygenation at a flow rate of 12 l.min−1 26. They reported a significantly less pronounced reduction in functional residual capacity in the high-flow group compared with the facemask group. Moreover, all patients in the former group maintained their arterial oxygen saturations for 10 min of apnoea, compared with no patients in the latter group 26. The optimal flow rate for effective pre-oxygenation and apnoeic oxygenation and ventilation has yet to be determined. However, what is clear from the evidence reported thus far is that as flow rates increase, so too does the efficiency of oxygenation in the apnoeic patient, as does the efficacy of pre-oxygenating the spontaneously ventilated patient. The higher the flow, the longer you go. There remain many unanswered questions in the field of apnoeic oxygenation. The underlying mechanisms of apnoeic oxygenation and ventilation in paediatric and obese patients remain unclear, as do the ideal flow rates of high-flow nasal oxygen delivery. Studies investigating patient-centred outcomes, such as satisfaction and quality of recovery scores, are also required. The adverse effects of hyperoxia and potential benefits of lower concentrations of oxygen delivered also remain an important avenue for investigation as there are no data in the peri-operative setting. Finally, data to assist in the prediction of safe apnoea times based on patient baseline characteristics would also be useful. O'Loughlin et al. have demonstrated that a low flow rate of oxygen administered directly into the trachea is associated with a prolonged safe apnoea time. These data further support the evidence we have highlighted that suggests efficiency of apnoeic oxygenation and ventilation can be achieved by using higher flow rates or lower delivery location. Risks of barotrauma must always be weighed against the benefits of apnoeic oxygenation techniques, but if this is considered, then: the lower you go, the lower the flow; the higher the flow, the longer you go. KE is an Editor of Anaesthesia and has received research, honoraria and educational funding from Fisher and Paykel Healthcare Ltd., GE Healthcare and Ambu. AP has received consulting and research support from Fisher and Paykel Healthcare Ltd. No other competing interests declared.