Influence of lung collapse distribution on the physiologic response to recruitment maneuvers during noninvasive continuous positive airway pressure

Noninvasive continuous positive airway pressure (n-CPAP) has been proposed for the treatment of hypoxemic acute respiratory failure (h-ARF). Recruitment maneuvers were shown to improve oxygenation, i.e., the ratio of arterial oxygen tension to inspiratory oxygen fraction (PaO2/FiO(2)), during either invasive mechanical ventilation, and n-CPAP, with a response depending on the distribution of lung collapse. We hypothesized that, during n-CPAP, early h-ARF patients with bilateral (B-L) distribution of lung involvement would benefit from recruitment maneuvers more than those with unilateral (U-L) involvement. To perform a recruitment maneuver, once a minute we increased the pressure applied to the airway from 10 cmH(2)O to 25 cmH(2)O for 8 s (SIGH). We enrolled 24 patients with h-ARF (12 B-L and 12 U-L) who underwent four consecutive trials: (1) 30 min breathing through a Venturi mask (V-MASK), (2) 1 h n-CPAP (n-CPAP(1)), (3) 1 h n-CPAP plus SIGH (n-CPAP(SIGH)), and (4) 1 h n-CPAP (n-CPAP(2)). Compared to V-MASK, n-CPAP at 10 cmH(2)O delivered via a helmet, increased PaO2/FiO(2) and decreased dyspnea in both B-L and U-L; furthermore, it reduced the respiratory rate and brought PaCO2 up to normal in B-L only. Compared to n-CPAP, n-CPAP(SIGH) significantly improved PaO2/FiO(2) in B-L (225 +/- A 88 vs. 308 +/- A 105, respectively), whereas it produced no further improvement in PaO2/FiO(2) in U-L (232 +/- A 72 vs. 231 +/- A 77, respectively). SIGH did not affect hemodynamics in both groups. Compared to n-CPAP, n-CPAP(SIGH) further improved arterial oxygenation in B-L patients, whereas it produced no additional benefit in those with U-L.