Selecting and adjusting PEEP

PEEP is the final phase variable in a breath cycle and describes the “baseline” pressure which remains behind after inspiratory pressure is released, as the patient exhales. Because of this, PEEP actually governs most of the respiratory cycle (of which traditionally 66% is expiratory time) and has the closest relationship to mean airway pressure out of all the phase variables. All ventilation could be viewed as footnotes to the application of PEEP. It is therefore all the more surprising that this chapter does not have any corresponding requirements to satisfy in 2023 CICM Primary Syllabus or in the CICM WCA document (“Ventilation”), where the trainees are expected to describe in detail the methods by which PEEP may be applied, but not which level to apply, or in which circumstances. That seems like something they should know. In order to address the authors’ own incomprehension of this topic, an entire chapter has been dedicated to the decision of how best to PEEP, and unto whom to do the PEEPing.

In summary:

  • The healthy lung should be ventilated with 5-8 cm H2O of PEEP
  • The ARDS lung will require higher PEEP (usually, over 12 cm H2O )
    • The methods of choosing the optimal PEEP in ARDS are discussed elsewhere
  • With bronchospasm, low PEEP or zero PEEP (ZEEP) is warranted
  • Negative end-expiratory pressure is to be avoided.

From the world of peer-reviewed literature, there are several excellent examples which have served to inform and organise the contents of this chapter. Of these, the vast majority seem to have been published by Luciano Gattinoni (Gattinoni et al, 2017; Gattinoni et al, 2015; Gatinnoni et al, 2010). Another excellent (but paywalled) reference is a 2016 review article by Bowton and Scott,  titled “Ventilatory management of the noninjured lung”.

Selecting a PEEP level for normal lungs

PEEP in healthy normal lungs is not viewed as an essential component of ventilation. However, some PEEP is probably still required. This seems like a non—controversial thing to modern ICU trainees, but it was not always so. When Pelosi et al (1999) were writing, there was “general agreement that the indiscriminate application of PEEP has no place in routine anesthesia in normal subjects”. Weirdly, the authors referenced this assertion with a 1990 article by Goran Hedenstierna, who actually recommended PEEP as a valid means of eliminating shunt, which – according to his sequantial CT data – developed within one minute of PEEP being discontinued. As a result of these sort of studies, the historical PEEPlessness has in the modern era become surpassed by the recognition of the importance of PEEP.

So, some PEEP is required. Exactly how much is a matter of some debate. There does not seem to be any specific studies of this, of if there are any, then they are tagged with such unique titles and abstracts that would defeat the usual search strategies used to find “PEEP AND select$ OR choos$”. Virtually everything turned up in the course of looking for this answer ends up being about ARDS. The half-page PEEP section from the Bersten chapter in Oh’s Manual (7th edition, p. 364) digresses extensively on ARDS, and mentions nothing about the bed-and-breakfast ventilator settings for the intubated drunk. 

Fortunately, there are other textbooks and authors. Kacmarek’s chapter for Egan’s Fundamentals of Respiratory Care (Ch. 48, p. 1078 from the 2016 edition) suggests a PEEP of 5-8 for most normal patients without offering a reference. Bowton & Scott (2016) agree with these numbers on the basis of a study by Manzano et al (2008), who ventilated normoxic patients with and without PEEP (5-8 cm H2O ), demonstrating a variety of benefits (better oxygenation, lower incidence of VAP) and no additional complications.

Selecting a PEEP level for poorly compliant lungs

This is explored in much greater detail in the Required Reading section for the fellowship exam. For the Part II, several questions have historically come up regarding how to optimise PEEP for “open lung ventilation”. In brief, there are about ten different strategies:

  • Use an arbitrarily high PEEP (15-20 cm H2O )
  • Use the ARDSNet PEEP/FiO2 escalation tables (setting the PEEP according to the severity of the oxygenation failure)
  • Titrate PEEP according to maximum compliance, i.e. set the PEEP which achieves the highest static compliance (this has the advantage of being tailored to each specific patient)
  • Set the PEEP to slightly above the lower inflection point of the pressure volume curve (therefore avoiding cyclic atelectasis)
  • Use a staircase recruitment (or derecruitment) manoeuvre to find the lowest PEEP at which the maximal oxygenation is maintained (this has the advantage of having a very pragmatic endpoint, SpO2)
  • Titrate PEEP to achieve the smallest intrapulmonary shunt using a PA catheter with continuous SvO2 monitoring
  • Titrate PEEP to achieve the lowest arterial minus end-tidal CO2 gradient (i.e. the PEEP at which dead space is minimal)
  • Use the transpulmonary pressure calculated from oesophageal balloon manometry, using oesophageal pressure (Pes) as a surrogate for pleural pressure (where TPP = Pplat - Pes). Adjust PEEP so that TPP at end-expiration is 0-10.
  • Using electrical impedance tomography, titrate PEEP to achieve the highest electrical impedance in the thorax (i.e. the greatest amount of aerated lung)
  • Use sequential CT scans to visually determine a PEEP at which the greatest volume of lung is recruited during end-expiration

Selecting a PEEP level for resistant airways

In short, keep it low. The main reasons for this are:

  • Avoid contributing to gas trapping
  • Avoid pneumothorax
  • Reduce haemodynamic compromise due to total PEEP (extrinsic + intrinsic)

Most authors recommend setting it below 5 cm H2O . Recommendations for the precise maximum PEEP in such patients are difficult to find, but generally one can assume that they would have something to do with the intrinsic PEEP level. Ranieri et al (1992) found that among COPD patients with gas trapping, setting a  PEEP higher than 85%  of the intrinsic PEEP resulted in more dynamic hyperinflation. Given the difficulties in measuring intrinsic PEEP, one may wish to empirically set a random low PEEP in these circumstances instead. Or, one may decide to use a PEEP of zero (ZEEP).


This ZEEP had once been the only EEP, so much so that contemporary articles about novel methods were titled “Ventilation with end-expiratory pressure in acute lung disease”, as opposed to without.

In the modern era, a zero PEEP is usually only set for patients with severe bronchospasm. In these people, the intrinsic PEEP generated in their lungs is already substantial, and anything contributed by the ventilator would only make things worse.

One other possible use of ZEEP is as a test of extubation readiness, i.e. if you can tolerate a PEEPless spontaneous breathing trial then you will probably be fine once off the ventilator.  Lakshminarayanan (2013) experimented with this and found it to be “safe and effective”. There was a substantial decrease in the re-intubation rate, which suggests that a) there was no significant derecruitment during the short period of ZEEP, and b) to pass a spontaneous breathing trial like this one requires the sort of muscular and cardiovascular reserve that promotes good post-extubation progress.

NEEP – Negative end expiratory pressure

At one stage, there was a misguided belief among physicians that active expiration was as important as active inspiration, and a negative pressure (as low as -4.0 cm H2O ) was applied to patients’ airways (Hill et al, 1965). All sorts of justifications defended this; for instance, Rogers et al (1981) found that it decreases ICP in brain injury. The practice fell into disuse in the late 1970s, because it was found to be harmful. Cyclical atelectatsis must cretainly be to blame. If the healthy mouse lung oozes cytokines after this (Cheng et al, 2002) then surely so does the lung of the critically ill patient.


Gattinoni, Luciano, et al. "Positive end-expiratory pressure: how to set it at the individual level." Annals of translational medicine 5.14 (2017).

Gattinoni, Luciano, Eleonora Carlesso, and Massimo Cressoni. "Selecting the ‘right’positive end-expiratory pressure level." Current opinion in critical care 21.1 (2015): 50-57.

Gattinoni, Luciano, et al. "Positive end-expiratory pressure." Current opinion in critical care 16.1 (2010): 39-44.

Dammann, J. Francis, and T. Crawford Mcaslan. "PEEP: its use in young patients with apparently normal lungs." Critical care medicine 7.1 (1979): 14-19.

Pelosi, Paolo, et al. "Positive end-expiratory pressure improves respiratory function in obese but not in normal subjects during anesthesia and paralysis." Anesthesiology: The Journal of the American Society of Anesthesiologists 91.5 (1999): 1221-1221.

Hedenstierna, G. "Gas exchange during anaesthesia." Acta Anaesthesiologica Scandinavica 34 (1990): 27-31.

Bowton, David L., and Louis Keith Scott. "Ventilatory Management of the Noninjured Lung." Clinics in chest medicine 37.4 (2016): 701-710.

Manzano, Francisco, et al. "Positive-end expiratory pressure reduces incidence of ventilator-associated pneumonia in nonhypoxemic patients." Critical care medicine 36.8 (2008): 2225-2231.

Leatherman, James. "Mechanical ventilation for severe asthma." Chest 147.6 (2015): 1671-1680.

Ranieri, V. Marco, et al. "Physiologic effects of positive end-expiratory pressure in patients with chronic obstructive pulmonary disease during acute ventilatory failure and controlled mechanical ventilation." American Review of Respiratory Disease 147.1 (1993): 5-13.

Caramez, Maria Paula, et al. "Paradoxical responses to positive end-expiratory pressure in patients with airway obstruction during controlled ventilation." Critical care medicine 33.7 (2005): 1519.

Falke, Konrad J., et al. "Ventilation with end-expiratory pressure in acute lung disease." The Journal of clinical investigation 51.9 (1972): 2315-2323.

Lakshminarayanan, Venkatesh, et al. "The Role Of Zero End Expiratory Pressure (zeep) In Extubation. A Prospective Study Investigating Zeep Without Pressure Support As A Method Of Successful Spontaneous Breathing Trials." B53. MECHANICAL VENTILATION. American Thoracic Society, 2013. A3027-A3027.

Rogers, Mark C., et al. "Negative End-expiratory Pressure (neep) Decreases Intracranial Pressure (icp)." Critical Care Medicine 9.3 (1981): 153.

HILL, J. DONALD, et al. "Correct use of respirator on cardiac patient after operation." Archives of Surgery 91.5 (1965): 775-778.

Cheng, Kuo-Chen, et al. "Ventilation with negative airway pressure induces a cytokine response in isolated mouse lung." Anesthesia & Analgesia 94.6 (2002): 1577-1582.