One-lung ventilation is required when providing anaesthesia for operative procedures in the thoracic cavity. During this process, hypoxaemia is reported to occur with an incidence of approximately 5–10%. 1 Unfortunately, hypoxaemia affects postoperative outcome as there is an increased risk of complications such as cognitive dysfunction, atrial fibrillation, renal failure, and pulmonary hypertension. 2 Thus, prevention and treatment of hypoxaemia associated with one-lung ventilation is a priority for anaesthetists. By enhancing our understanding of the airway, ventilation, and perfusion, this editorial explores how the results of recent studies may refine future clinical practice. Previously, considerations of hypoxaemia associated with one-lung ventilation have related to airway complications. In the past 50 yr, there have been innovations in the design of airway devices that enable correct lung isolation, ventilation, and suction. 3 Fibreoptic bronchoscopy to check precise positioning of double-lumen endobronchial tubes has been recommended to enhance the safety of one-lung ventilation. 4 More recently, data show that malpositioning of double-lumen endobronchial tubes is still a problem, with an incidence of more than 32%, for anaesthetists with limited experience of thoracic anaesthesia. 5 Analysis of automated anaesthetic records has provided some clarification concerning the perceived difficulty of right-sided doublelumen endobronchial tubes which have a slotted cuff. In an observational study, fibreoptic bronchoscopy was used routinely, and no association was found between the side of double-lumen endobronchial tube and hypoxaemia during one-lung ventilation. 6

In addition to airway considerations, ventilation strategies in relation to hypoxaemia have been studied. From past data, we understand that the application of positive end-expiratory pressure may reduce atelectasis, decrease intrapulmonary shunt, and improve oxygenation during one-lung anaesthesia. 7 Whilst this technique may be beneficial to the atelectatic lung after a recruitment manoeuvre, it may to be harmful in patients with hyperinflated lungs as these reside above the lower inflection point of the compliance graph. 8 In addition, data from a recent randomized controlled trial (RCT) show that mode of ventilation, that is to say, volume or pressure control, does not appear to affect oxygenation during one-lung ventilation. 9 Our understanding of the relationship between one-lung ventilation and pathogenesis of hypoxaemia is improving as a result of studies evaluating lung inflammation. For instance, compared with a low tidal volume of 5 ml kgJ1, one-lung ventilation at a high tidal volume of 10 ml kgJ1 has been shown to be associated with a significantly augmented inflammatory response. 10 With an increased duration of exposure to one-lung ventilation, there is production of inflammatory mediators, recruitment of neutrophils, and possible damage to the alveolar capillary membrane. 11 Fortunately, data from two recent RCTs suggest that this inflammatory process may be attenuated by inhalational agents. In one of them, alveolar granulocytes, IL-8, and sICAM-1 (soluble intercellular adhesion molecule) in fluid obtained by lavage of the ventilated lung were significantly lower in patients receiving desflurane than in those who had propofol. 12 These results are supported by another RCT