Low-Tidal-Volume Ventilation and Mortality in Patients With Acute Brain Injury: A Secondary Analysis of an International Observational Study.

Abstract

Low-tidal-volume ventilation (LTVV) improves outcomes in critically ill patients, but its impact in patients with acute brain injuries (ABIs) is less certain. What is the association between LTVV and mortality in mechanically ventilated patients with ABI? We did a secondary analysis of a prospective observational study (NCT03400904; https://clinicaltrials.gov/study/NCT03400904). We compared LTVV (≤ 8 mL/kg predicted body weight [PBW]) with tidal volumes > 8 mL/kg PBW over the first 7 days of mechanical ventilation. Alternate analyses used lower thresholds for LTVV. Marginal structural Cox models were used to evaluate the association between LTVV and ICU mortality up to 60 days. Stabilized inverse probability treatment and censoring weights were generated using multivariable logistic regression adjusted for baseline and time-dependent confounders. A total of 1,510 patients from 73 ICUs across 18 countries were included. The mean age was 52 years, 513 patients (34.0%) were female, and the most common ABI etiology was traumatic brain injury (n = 726; 48.1%). ARDS developed in 137 patients (9.2%). In patients receiving LTVV, adjusted incidence of ICU mortality was 40.2% (95% CI, 19.2%-61.1%), vs 59.7% (95% CI, 44.0%-75.4%) in patients receiving tidal volumes > 8 mL/kg PBW (marginal hazard ratio, 0.54; 95% CI, 0.33-0.88). There was no heterogeneity of treatment effect in subgroup analyses, and sensitivity analyses for unmeasured confounding yielded similar results. However, associations were less clear at lower thresholds of LTVV. In this predominantly non-ARDS cohort of patients with ABI, LTVV over the first 7 days of mechanical ventilation was associated with lower ICU mortality up to 60 days, vs tidal volumes > 8 mL/kg PBW. Future research should investigate effects in patients with ABI and ARDS, use of lower LTVV thresholds, and impact on additional end points including functional outcomes and adverse events.