Scientific Quality of Experimental Methods.

*Details obtained from a separate publication of a subgroup with trauma-induced ARDS by McKinley et al (n = 67)[34].PVL = pressure and volume-limitation; LIS = lung injury score[21]; APACHE II = Acute Physiology and Chronic Health Evaluation II[22]; SAPS = simplified acute physiology score[47]; CCS = clinical classification score[48]; MODS = multiple organ dysfunction score[49]; APACHE III = Acute Physiology and Chronic Health Evaluation[50]; ISS = illness severity score[51]; NA = not available.Assessment:aBaseline similarity: Factors assessed at the time of randomization included Line 1: age; Line 2: severity of pulmonary injury (PaO2/FiO2 or LIS[21] or A-a gradient or oxygen index or compliance); Line 3: illness severity (APACHE II[22] or SAPS[47] or MODS[49] or APACHE III[50] or ISS[51] and Line 4: sepsis. Each variable was assessed as similar between treatment groups or favoring PVL or control.bExperimental cointerventions: Line 1: We assessed for the use of corticosteroids, inhaled nitric oxide, prone positioning, high frequency oscillation, extracorporeal circulation and surfactant; Line 2: We described the frequency with which experimental cointerventions were utilized in each study and between treatment groups within studies.cSedation: We characterized sedation management as guided by an explicit protocol, suggested guideline or at the discretion of clinicians.dWeaning: We characterized weaning management as guided by an explicit protocol, suggested guideline or at the discretion of clinicians.eEarly stopping: Line 1: We assessed for early trial termination; Line 2: We assessed whether the study stopped early for benefit or futility and Line 3We assessed for explicit a priori stopping rules.