Clinicians' approaches to mechanical ventilation in acute lung injury and ARDS

Study objectives: To examine clinicians' approaches to mechanical ventilation in patients with acute lung injury (ALI; Pao(2)/fraction of inspired oxygen [FIO2] less than or equal to 300) and compare ventilator settings in patients with ARDS (Pao(2)/FIO2 less than or equal to 200) to settings in patients with milder oxygenation impairment (Pao(2)/FIO2 of 201 to 300). Design: Retrospective analysis of baseline data from prospective randomized trials conducted by the National Institutes of Health ARDS Network between 1996 and 1999. Setting: Ten clinical centers comprising 24 hospitals and 74 medical and surgical ICUs of the ARDS Network. Measurements and results: The most common mode of mechanical ventilation in both groups was volume-assist control (56%). Synchronized intermittent mandatory ventilation (SIMV) or SIMV with pressure support was used more often in patients with Pao(2)/FIO2 of 201 to 300 than in patients with ARDS. The use of pressure-control ventilation was uncommon (10% overall), as was the use of permissive hypercapnia (6% of patients with ARDS and 3% of patients with Pao(2)/FIO2 of 201 to 300). The mean +/- SD tidal volume was 10.3 +/- 2 mL/kg of predicted body weight or 8.6 +/- 2 mL/kg of measured weight for patients with ARDS, and was not significantly different for patients with Pao(2)/FIO2 of 201 to 300. Plateau pressures (Pplats) were lower in the Pao(2)/FIO2 of 201 to 300 group (27 +/- 7 vs 31 +/- 8 for the ARDS group; p = 0.0003) and were > 35 cm H2O in 26% of patients. Seventy-eight percent of patients with ARDS received less than or equal to 10 cm H2O of positive end-expiratory pressure. Conclusions: Physicians in ARDS Network centers caring for patients early in the course of ALI/ARDS used volume-targeted ventilation and selected tidal volumes that resulted in Pplats generally < 35 cm H2O. The average tidal volume was similar for patients with ARDS vs those with milder oxygenation deficits.