Critical Care (Jun 2022)

The PROMIZING trial enrollment algorithm for early identification of patients ready for unassisted breathing

  • Clement Brault,
  • Jordi Mancebo,
  • Juan-Carlos Suarez Montero,
  • Tracey Bentall,
  • Karen E. A. Burns,
  • Thomas Piraino,
  • François Lellouche,
  • Pierre-Alexandre Bouchard,
  • Emmanuel Charbonney,
  • Guillaume Carteaux,
  • Tommaso Maraffi,
  • Gaëtan Beduneau,
  • Alain Mercat,
  • Yoanna Skrobik,
  • Fei Zuo,
  • Myriam Lafreniere-Roula,
  • Kevin Thorpe,
  • Laurent Brochard,
  • Karen J. Bosma

DOI
https://doi.org/10.1186/s13054-022-04063-4
Journal volume & issue
Vol. 26, no. 1
pp. 1 – 9

Abstract

Read online

Abstract Background Liberating patients from mechanical ventilation (MV) requires a systematic approach. In the context of a clinical trial, we developed a simple algorithm to identify patients who tolerate assisted ventilation but still require ongoing MV to be randomized. We report on the use of this algorithm to screen potential trial participants for enrollment and subsequent randomization in the Proportional Assist Ventilation for Minimizing the Duration of MV (PROMIZING) study. Methods The algorithm included five steps: enrollment criteria, pressure support ventilation (PSV) tolerance trial, weaning criteria, continuous positive airway pressure (CPAP) tolerance trial (0 cmH2O during 2 min) and spontaneous breathing trial (SBT): on fraction of inspired oxygen (FiO2) 40% for 30–120 min. Patients who failed the weaning criteria, CPAP Zero trial, or SBT were randomized. We describe the characteristics of patients who were initially enrolled, but passed all steps in the algorithm and consequently were not randomized. Results Among the 374 enrolled patients, 93 (25%) patients passed all five steps. At time of enrollment, most patients were on PSV (87%) with a mean (± standard deviation) FiO2 of 34 (± 6) %, PSV of 8.7 (± 2.9) cmH2O, and positive end-expiratory pressure of 6.1 (± 1.6) cmH2O. Minute ventilation was 9.0 (± 3.1) L/min with a respiratory rate of 17.4 (± 4.4) breaths/min. Patients were liberated from MV with a median [interquartile range] delay between initial screening and extubation of 5 [1–49] hours. Only 7 (8%) patients required reintubation. Conclusion The trial algorithm permitted identification of 93 (25%) patients who were ready to extubate, while their clinicians predicted a duration of ventilation higher than 24 h.

Keywords