Interpreting individualised positive end‐expiratory pressure

Wang et al. should be congratulated for conducting an assessor-blind randomised trial in older patients undergoing lung cancer surgery [1]. Their study also looked at whether electrical impedance tomography-guided individualised positive end-expiratory pressure (PEEP) could improve postoperative outcomes. The intervention lowered driving pressure and improved oxygenation, but it did not reduce the reported composite postoperative pulmonary complication endpoint. We agree that better intra-operative mechanics should not be equated with better postoperative recovery. However, these findings should be interpreted as showing no benefit for the protocol and endpoint tested, rather than as definitive evidence against individualised PEEP in thoracic surgery. A multicentre trial of driving pressure-guided ventilation improved respiratory mechanics in a similar manner without reducing postoperative pulmonary complications, whereas pooled one-lung ventilation evidence remains heterogeneous and dependent on the titration method; comparator PEEP; and outcome definitions [2, 3]. One reason may be the composition of the primary outcome. In the study by Wang et al., prolonged air leak was the most frequent pulmonary complication, occurring in 13% of patients in each group. This outcome is important after lung resection, but it is influenced by surgical and parenchymal factors, including lung function and extent of resection, and may therefore be less amenable to modification by a short ventilatory intervention than atelectasis or oxygenation-related respiratory failure [4]. When events with different severity and biological responsiveness are collapsed into a single endpoint, an effect on ventilation-sensitive complications may be diluted. The intervention tested was a two-point electrical impedance tomography-guided PEEP strategy limited to intra-operative one-lung ventilation and immediate postoperative two-lung ventilation, in a single-centre cohort that did not include patients with a BMI > 35 kg.m-2 and chronic airway disease. Recruitment manoeuvres were used in both groups, and PEEP was not titrated during major intra-operative physiological changes [1]. This differs from iPROVE one-lung ventilation, in which an individualised peri-operative open-lung bundle extending into postoperative respiratory support reduced severe postoperative pulmonary complications [5]. Thus, the neutral primary outcome may define the limits of electrical impedance tomography-guided PEEP titration alone, rather than rule out bundled strategies or phenotype-selected subgroups. This trial discourages expecting a large reduction in a broad thoracic postoperative pulmonary complication composite from electrical impedance tomography-guided PEEP titration alone in similar patients who are older, non-obese and without chronic airway disease. It also raises useful questions about which components were plausibly modifiable by ventilation; whether patients with more recruitable lungs or higher driving pressures were most likely to benefit; and whether outcomes varied according to the achieved change in driving pressure or the type of lung resection. Presenting these analyses, even as exploratory, would sharpen the clinical interpretation of this important trial.