Lung recruitment and hyperinflation with higher positive end-expiratory pressure in COVID-19: a computed tomography study

Abstract Background A higher (> 10 cmH 2 O) positive end-expiratory pressure (PEEP) is commonly used in patients with moderate to severe hypoxemia due to the novel coronavirus disease (COVID-19). However, compliance more commonly decreases when PEEP is increased from 10 to 15 cmH 2 O, as for lung hyperinflation. In this study, we directly measured lung recruitment and hyperinflation induced by increasing PEEP from 10 to 15 cmH 2 O in mechanically ventilated patients with COVID-19. Methods Twenty mechanically ventilated patients with COVID-19 underwent a lung computed tomography (CT) at 10 and 15 cmH 2 O of airway pressure. Gas exchange and compliance were then measured with 10 and 15 cmH 2 O of PEEP. Recruitment was computed as the decrease of the non-aerated lung volume (density above − 100 HU) and hyperinflation as the increase of the over-aerated lung volume (density below − 900 HU). If recruitment was larger than hyperinflation, the net morphological response was “recruitment”; otherwise, it was “hyperinflation”. Results With 10 cmH 2 O of PEEP, the median (Q1-Q3) arterial tension to the inspiratory fraction of oxygen (PaO 2 :FiO 2 ) was 146 (107–197) mmHg. The net morphological response to increasing PEEP was recruitment in nine (45%) patients and hyperinflation in eleven (55%). Oxygenation improved in twelve (60%) patients, but compliance in only three (15%). Compliance with 10 cmH 2 O of PEEP ≤ 45 cmH 2 O/ml predicted net recruitment in response to increasing PEEP with a positive predictive value of 0.89 (95%-confidence intervals: 0.49-1.00) and a negative predictive value of 0.91 (0.59-1.00). Conclusions The morphological response to a higher PEEP is largely variable in patients with COVID-19, ranging from net recruitment (in half of the patients) to net hyperinflation (in the other half). Baseline compliance may help to predict the individual response to increasing PEEP.