For ARDS, esophageal pressure-guided PEEP disappoints in EPVent-2 trial

  • Beitler JR & al.
  • JAMA
  • 18 Feb 2019

  • curated by Jenny Blair, MD
  • Clinical Essentials
Access to the full content of this site is available only to registered healthcare professionals. Access to the full content of this site is available only to registered healthcare professionals.


  • For patients with acute respiratory distress syndrome (ARDS), positive end-expiratory pressure (PEEP) titration guided by esophageal pressure (PES) offers no clear advantage over empirical high PEEP-FIO2.
  • Editorial discusses differences in protocol, results vs EPVent and concludes: “for most patients with moderate to severe ARDS, the use of a simple and widely available tool to adjust PEEP will result in similar clinical outcomes to those of a more invasive, costly, and cumbersome strategy.”

Why this matters

  • Benefit of PEEP optimization is unclear, and the process is difficult because it must not only improve oxygenation but also prevent ventilator-induced lung injury (VILI). 
  • Pleural pressure correlates with PES (measured by esophageal manometer).
  • PES-guided PEEP appeared advantageous in 2008 EPVent trial.

Key results

  • PES-guided vs empirical PEEP:
    • Primary outcome: no significant difference.
    • Death: no significant difference.
    • Days free from mechanical ventilation: no significant difference.

Study design

  • 14-hospital randomized clinical EPVent-2 (n=200).
  • Patients aged ≥16 years with moderate to severe ARDS were randomly assigned to receive PES-guided PEEP vs empirical high PEEP-FIO2.
  • Outcome: ranked composite score incorporating death and days free from mechanical ventilation among survivors through day 28.
  • Funding: NIH.


  • Underpowered for meaningful survival differences. 
  • Prone positioning prohibited except as rescue.

Please confirm your acceptance

To gain full access to GPnotebook please confirm:

By submitting here you confirm that you have accepted Terms of Use and Privacy Policy of GPnotebook.