Objective: To determine the usual practice for setting tidal volume and other ventilatory parameters in patients with acute respiratory distress syndrome (ARDS) in the late 1990s and to determine the independent effects of these practices on intensive care unit mortality.
Design: Subanalysis of a prospective observational study. Multivariable logistic regression was used to analyze the effects of ventilatory management on mortality.
Setting: A total of 361 intensive care units in 20 countries in March 1998.
Patients: A total of 467 mechanically ventilated patients with ARDS.
Measurements and main results: The mean tidal volume used in the first week of ARDS was 8.8 mL/kg measured body weight, and there was great variability in these tidal volumes (sd = 2.0). Tidal volumes were significantly lower in patients with (n = 265) than without (n = 202) a recorded plateau pressure (8.6 vs. 9.1 mL/kg, p = .01). The overall intensive care unit mortality rate was 60.2%. In addition to the strong influence of organ failures and higher levels of inspired oxygen, late-onset ARDS (onset after >48 hrs of mechanical ventilation; odds ratio, 2.09) was independently associated with mortality. In addition, lower levels of positive end-expiratory pressure were independently associated with higher mortality (odds ratio, 0.91; 1 cm of H2O increments). Neither inspiratory pressures nor tidal volumes were independently associated with mortality, and there was no evidence of increased mortality with the use of lower inspiratory pressures.
Conclusions: This descriptive study demonstrated considerable interpatient variability in tidal volumes during the study period. In addition to traditional prognostic indicators, timing of ARDS onset and the use of low levels of positive end-expiratory pressure or no positive end-expiratory pressure during the first week may adversely influence outcome in ARDS patients.