Should we automate PEEP ?

Before we dig in this tough topic, let’s agree or disagree that that there is nothing called optimal or best PEEP. Regardless of the PEEP level used, there will be overdistention in some areas, some under distention and collapse in other areas of the lungs.

Here are some points to keep in mind when we talk PEEP

  • The normal and the diseased lungs are markedly heterogenic with each lobule or even each of the approximately 500 million alveoli have their own mechanics
  • The pleural pressures are not uniform and thus the trans-alveolar pressures will vary depending on the location of the alveoli
  • Not all Lungs especially ARDS are recruitable
  • “Not all ARDS are the same” but studies don’t differentiate ARDS phenotypes
  • “PEEP does not recruit, rather prevents derecruitment”
  • We don’t have much agreements in the literature on how to set PEEP or if PEEP affect mortality
  • The respiratory mechanics change frequently, even between breaths to breaths but we don’t change the PEEP frequently enough, so what was good 10 minutes ago might not be good now

Why automate PEEP or let the machine adjust PEEP?

Simple answer, because we honestly don’t know how despite 50 years of research, and we can’t do it continuously or be at the bedside all the time while the respiratory mechanics are continuously changing

What are the benefits?

  • To minimize the over and under distention of one level of PEEP
  • The ventilator can be consistent on the way it adjust the PEEP per specific algorithms

Are there any modes that currently automate PEEP?

Yes, INTELLiVENT‑ASV from Hamilton was the first mode to automate PEEP, among all other parameters of ventilation and proved to be an effective mode of ventilation from intubation to extubation. However the algorithms used is the ARDSnet PEEP-FiO2 table, though this table have been used for the last 2 decades, it is very non physiological and does not take in account the lung recruitability or if the higher PEEP could be beneficial or harmful.

PMLV (Programmed Multi Level Ventilation) uses alternating 2 or 3 levels of PEEP but they are set by the clinicians

How can the ventilator choose PEEP level ?

As mentioned above, there is no agreement in the literature on the best method of setting PEEP. There are so many different physiologic ways, lets name some:

  • Pressure Volume curve (Hysteresis of the curve, Lower inflection point of the inspiratory limb, point of maximal curvature on the expiratory limb)
  • Incremental or decremental PEEP trial
  • Best compliance and lowest driving / tidal pressure
  • Expiratory time constant during different PEEP levels
  • Esophageal balloon monitoring and transpulmonary pressures
  • Volumetric capnometry (Dead space and VCO2)
  • Electrical Impedance Tomography (EIT) signals of over and under inflation during different PEEP levels
  • According to measured FRC

Now, the next question is: can the ventilator do those maneuvers by itself? when ? and how often?

Currently most of those maneuvers, signals, information (except EIT, though the ventilator can get the signal from EIT monitor) are measured by the new generation ventilators. The ventilator would know when respiratory compliance changes (it measures it breath by breath)

As a computer, the ventilator can be programed to do any of those maneuvers independently, and at a programmed intervals according to a specific algorithm and through feedback system can change the current PEEP settings up or down.

Goes without saying, that this would be controversial and needs more studying and research, but for now its just a blog with some ideas

5 Replies to “Should we automate PEEP ?”

  1. Autorizar la peep por medio de ventiladores inteligentes , será otra opción más en beneficio del paciente , con la idea es siempre la individualización y monitoreo dinámico a cada paciente .

  2. First, I agree with the premise that regardless of how you set it, lung heterogeneity prevents any sort of uniform or universal respiratory system distention. I don’t agree with abandoning the term ‘optimal’ though. Define it differently yes, but we have to have some point(s) we determine to have the best outcome.

    Automating the process still needs human intervention and decision making. Also, the algorithms should depend on the disease state and past history. Example, Covid pneumonia in a COPD lung has to be treated differently than in a previously healthy lung due to those differences In heterogeneity. So choosing best compliance parameters for the healthy lung may not work for COPD lungs.

    I’d say for starters, using the P/V loop and choosing places between the inflection points and let the vent do the work MIGHT be an idea. Or better, making several calculations of FRC and find the PEEP that maintains that volume (+/- some acceptable value) during different body positions and/or changes in compliance.

    That said, the goal of preventing de-recruitment may not be achieved by frequent changes in PEEP. Regardless of which flawed model we use for gas exchange and pulmonary architecture, it does take time to re-recruit changes in each lung unit. So, if left to constantly or frequently changing PEEP we never achieve best/optimal recruitment.

    My initial thoughts any way.

  3. Los respiradores de VMNI de gama media -alta ya permiten un modo de ventilación con PEEP variable desde hace años que resulta muy útil en determinados tipos de enfermos, principalmente en obesos con variabilidad de la necesidad dé CPAP por los cambios posturales y en pacientes con EPOC grave e importante atrapamiento aéreo

    1. You’re right, but I was thinking something more advanced the way I interpreted the question. I think your comments might give us a better starting point than PMLV describe previously.

  4. Rather than having the ventilator set an “Optimal PEEP” (which may not be attainable) is there some benefit to having the ventilator provide “Noisy” PEEP instead? Could the variable End-Expiratory Lung Volume help address the heterogeneity of PEEP distribution?

Leave a Reply

Your email address will not be published. Required fields are marked *

Share on Social Media
Follow us on Social Media