Examining Tidal Volume Practices in Clinical Trials of ALI/ARDS

Clinical Center

Investigators

Linked publications & trials

Abstract

Following the ARDSNet report of the survival benefit of a 6 vs.12 mL/kg predicted body weight (PBW) Tidal Volume (TV) protocol (ARMA, 2000) for Acute Respiratory Distress Syndrome/Acute Lung Injury (ARDS/ALI), the importance of protective lung ventilation has become widely accepted. However, questions exist regarding how low TVs should be. We, therefore, investigated TV practice at centers investigating ALI/ARDS since publication of ARMA. Databases were searched using relevant terms. Randomized clinical trials (RCTs) were included if they studied non-TV-related therapies for ALI/ARDS and reported TVs. Extracted data included enrollment dates, patient age, the earliest reported TV (mL/kg-PBW designated below as mL), PaO2/FiO2, plateau pressure, whether trials were from ARDSNet vs. non-ARDSNet centers, and if an ARDSNet low-TV protocol was used following randomization. Twelve RCTs with start of enrollment dates ranging from June 2000 to August 2007 met inclusion criteria. Eleven trials reported TVs based on PBW and for one a conversion factor for total body weight (+1.55mL) was used. Three RCTs were from ARDSNet (75-1000 patients) and nine from non-ARDSNet (16-983 patients) centers. The ARDSNet low-TV protocol was used after randomization in 10 trials. Differences in pre-randomization TVs among trials could not be explained by baseline PaO2/FiO2, plateau pressures, or patient age (all p=NS). Overall, the average pre-randomization TV was 8.23 +/-0.32mL in the year 2000, decreased on average 0.17 +/- 0.08mL per year (p=0.04) and was lower in ARDSNet versus non-ARDSNet centers (7.09 +/- 0.35mL vs. 7.88 +/- 0.21mL; p=0.05). In a multivariate analysis including enrollment year and center type, estimated TVs for ARDSNet and non-ARDSNet centers in 2000 were 7.66 +/- 0.43mL vs. 8.32 +/- 0.29mL respectively (p=0.076) and decreased similarly over time (0.14 +/- 0.07mL per year overall, p=0.05). After randomization, TVs decreased significantly in ARDSNet and non-ARDSNet centers employing an ARDSNet low-TV protocol (-0.80 +/- 0.18mL and -1.08 +/- 0.13mL respectively when examined at 2003, the median year studied; both p<0.0001) but not at centers not using this protocol (-0.03 +/- 0.25mL; p=0.89). These changes differed significantly comparing centers using vs. not using the ARDSNet protocol (p0.01) but not when comparing ARDSNet vs. non-ARDSNet centers using this protocol (p=NS). Routine TVs at centers investigating ALI/ARDS have gradually decreased since 2000 and tended to be lower at ARDSNet vs. non-ARDSNet centers. Following randomization, patients in trials employing the ARDSNet low-TV protocol had their TVs significantly reduced. Whether using TVs different from routine practice influences detection of potential beneficial or harmful effects in RCTs of non-TV-related therapies for ARDS/ALI is unknown, but could confound extrapolation of such findings clinically. This study has been expanded to include observational trials of ALI and ARDS and a manuscript is in preparation. Findings from the study were presented at the 2012 ATS International Conference.