By Richard Kallet, MS, RRT, FCCM

Director of Quality Assurance, Respiratory Care Services, Department of Anesthesia, San Francisco General Hospital

Mr. Kallet reports he is a major stockholder of and a business advisory board member for the Asthma & Allergy Prevention Company and receives grant/research support from Nihon-Kohden.

Fifty years ago, positive end-expiratory pressure (PEEP) was introduced as an effective technique for improving oxygenation in patients with large intrapulmonary shunt, the hallmark of acute respiratory distress syndrome (ARDS).1 Although PEEP remains the primary means for stabilizing oxygenation in ARDS, consensus on how to approach setting it remains elusive.2 This is a narrative review on how our understanding and approach to PEEP has evolved over the past half century.


Functional residual capacity (FRC) is essentially the alveolar volume and, as such, the primary determinant of both oxygenation and respiratory system compliance (CRS).3 Historically, the impact of PEEP has been assessed according to its effect on oxygenation or CRS as signifiers of changes in FRC. This is a matter of bedside expedience understood implicitly but rarely stated explicitly. A historically accurate statement might read: The primary impact of PEEP is the stabilization of underinflated alveoli and the recruitment of collapsed small airways and alveoli, thereby increasing FRC and CRS while reducing intrapulmonary shunt so as to allow mechanical ventilation at a relatively safer inspired oxygen fraction (FiO2; i.e., < 0.70).4-6

PEEP also produces lung-protective effects. Repetitive opening and closing of small airways and alveoli causes shearing of both the airway epithelium (from breaking and displacing liquid bridges/plugs) and the alveolar epithelium (from asymmetrical stress/strain development between patent alveoli adjacent to collapsed but recruitable alveoli).7 Restoring FRC toward normal is lung-protective in that it: 1) reduces the formation of liquid bridges and promotes alveolar edema fluid translocation from airspaces to the interstitium, 2) increases alveolar surface area to accommodate tidal ventilation (thus, reducing excessive tidal strain), and 3) allows ventilation at relatively nontoxic FiO2. Particularly noteworthy in this regard is emerging evidence that hyperoxia appears to potentiate stretch-related injury.8


PEEP does not “recruit” collapsed alveoli but rather stabilizes partially inflated alveoli by counteracting forces that promote collapse, such as increased alveolar surface tension and superimposed hydrostatic forces, and, therefore, prevents tidal “de-recruitment” in mid-level and some dependent regions. On the other hand, recruitment is an inspiratory phenomenon. Its goal is to achieve a threshold opening pressure sufficient to overcome both retractile and compressive forces as well as rupturing liquid bridges and displacing liquid plugs that obstruct the peripheral airspaces.7,9

The clinical surrogate for threshold opening pressure is plateau pressure (Pplat), which represents the mean, quasi-static alveolar pressure at end inspiration. The confusion over the role of PEEP in recruitment stems from the fact that PEEP almost invariably increases Pplat. This was particularly true when a traditional tidal volume (VT) of 12-15 mL/kg was used, often producing a Pplat ranging from 35-50 cm H2O. In ARDS, threshold opening pressures follow a bimodal distribution. Whereas most lung units achieve full recruitment with a Pplat of 20-35 cm H2O, in severe ARDS, dorsal-caudal regions exposed to substantial compressive/occlusive forces require opening pressures of 40-60 cm H2O.9 This explains why moderate levels of PEEP usually are sufficient to stabilize oxygenation in most ARDS cases.

The historical reluctance to use high PEEP levels stems from a misinterpretation of the foundational studies conducted in the 1970s. Without question, high PEEP levels can cause hemodynamic compromise, reduced systemic oxygen delivery, and barotrauma. Yet, these studies were performed using a VT of 10-15 mL/kg or higher. This often resulted in extraordinarily high alveolar pressures. It also increased the time needed for pressure dissipation during the expiratory phase (i.e., the “amplitude constant” that governs pressure-volume equilibration in the lungs),10 further impeding venous return and right ventricular output.

The conclusions drawn from these studies were that setting PEEP above 10 cm H2O increased the risk of hemodynamic instability and barotrauma, and that levels > 15 cm H2O should be avoided.3 In this context, it was particularly unfortunate that a crucial PEEP study from 1978 was essentially ignored. Suter et al found that CRS continued to improve even at a PEEP of 15 cm H2O when a physiologic VT (5-7 mL/kg) was used.11 Therefore, the negative effects of PEEP were VT-dependent and largely avoidable. These findings from 1978 presaged lung protective ventilation practices in the early 2000s. What essentially escaped the focus of the pulmonary critical care profession for decades was a sense of historical reflection. Specifically, the use of supra-physiologic VT predated the advent of PEEP, and although initially used for treating postoperative atelectasis, it quickly was incorporated into standard mechanical ventilation practice. Even after its introduction in 1967, it took another decade before the relative contributions of PEEP and VT were investigated. In practical terms, from 1967 until publication of the seminal NIH ARDSNet ARMA trial in 2000,12 mechanical ventilation practices generally relied on a supra-physiologic VT with relatively low PEEP and high FiO2 to manage ARDS, with sobering results in terms of iatrogenic lung injury.


Since the early 1970s, divergent PEEP strategies typically focused on “optimizing” either oxygenation or CRS. Another approach that gained widespread acceptance was using the “least PEEP” necessary to ensure adequate arterial oxygen tension (e.g., partial pressure of oxygen, PaO2, of approximately 70 mmHg) while avoiding excessive hyperoxia, barotrauma, and hemodynamic compromise.13 Although these words evoke advocacy for low PEEP, it essentially argued against using PEEP to “optimize” oxygenation as was advocated by “super-PEEP” adherents at that time.14 In essence, the FiO2/PEEP table used in the NIH ARDSNet ARMA study was largely consistent with the least PEEP philosophy, as it attempted to balance the benefits and risks of PEEP with the risks of hyperoxia while maintaining a reasonable PaO2 (55-80 mmHg).12 In the early, exudative phase of ARDS when the lungs typically are most amenable to recruitment, ARDSNet-guided PEEP is increased aggressively to stabilize oxygenation, then just as aggressively titrated downward to find the minimum PEEP needed to maintain modest oxygenation goals. In the early 1990s, the “open lung ventilation” (OLV) strategy heralded a shift in focus toward lung protection. The first iteration advocated briefly recruiting the lungs with pressure ventilation at 55 cm H2O and PEEP of 16 cm H2O. This was followed by reducing tidal driving pressure to < 20 cm H2O to prevent stretch-related injury and inverse-ratio ventilation titrated to an intrinsic PEEP of 16 cm H2O to prevent sheer-related injury.15 Since then, OLV had been modified to incorporate traditional inspiratory:expiratory ratios with low VT and external PEEP. PEEP is titrated to prevent de-recruitment by using a decremental trial that adjusts PEEP to 2 cm H2O above the level when deterioration in either oxygenation or CRS becomes apparent.

However, when reviewing OLV studies, one discovers that the mean optimal PEEP was only 10-12 cm H2O or had resulted only in modest reductions in mean PEEP from approximately 12 to 9 cm H2O. This raises the question of whether this approach offers any advantage over incrementally upward PEEP adjustments. For context, large clinical trials of ARDS that combined higher PEEP with low VT using either oxygenation- or mechanics-based titration protocols consistently reported day 1 mean PEEP levels of approximately 14 to 16 cm H2O.16-18 Unfortunately, a recent large, multicenter, randomized, controlled trial of OLV reported a higher mortality with OLV compared to the ARDSNet ARMA strategy.19 Although important unresolved questions about this study remain, its impact is likely that OLV will be used only as rescue therapy.

A newer approach has been titrating PEEP to maintain an end-expiratory transpulmonary pressure between 0 and 10 cm H2O, where transpulmonary pressure is equal to PEEP minus esophageal pressure at end expiration. The authors of a pilot study found that the transpulmonary pressure-guided strategy resulted in both a higher mean PEEP and mean PaO2 /FiO2 (18 cm H2O and 280 mmHg, respectively) compared to the ARDSNet ARMA strategy (12 cm H2O and 191 mmHg, respectively), both of which were achieved at an FiO2 < 0.60.20 From a practical standpoint, this invasive technique is not necessary to manage most ARDS cases, but might be useful in managing cases complicated by morbid obesity or abdominal compartment syndrome. However, even in these cases, PEEP can be titrated empirically, measuring intra-abdominal pressure as a guide. Other evidence garnered from the literature also may help answer the following question: After a half century of treating ARDS with PEEP, what range typically is needed in most cases? The authors of early physiologic studies speculated that by improving FRC and CRS, PEEP moved tidal ventilation onto the steep portion of the inflation pressure-volume (P-V) curve (i.e., above the lower inflection point, or LIP).4,5 This concept was supported by other studies, which found that found setting PEEP 3 cm H2O above LIP markedly improved oxygenation and that the average LIP was approximately 9 ± 3 cm H2O, or a set PEEP of approximately 12 cm H2O.21,22 Data culled from 16 trials totaling 197 discreet LIP measurements revealed a median LIP of 10 (interquartile range of 8 to 3) cm H2O, which translates into PEEP settings between 11 and 16 cm H2O.23 The authors of that review also found six additional studies that reported mean values of LIP between 8 and 11 cm H2O with a subsequent corresponding PEEP of 11-14 cm H2O.

Setting PEEP to prevent alveolar collapse also has been advocated based on the findings from CT studies. Increased dorsal lung densities are believed to represent compressive atelectasis from the weight of the overlying edematous lung. Setting a minimal PEEP of 11-14 cm H2O was proposed to keep the dependent lung zones open at end expiration.24 This was based on the sternovertebral height of a supine adult (12-25 cm) and the average tissue density in ARDS of 0.7 gm/cm3, which produces a PEEP range of 8-18 cm H2O. It’s particularly noteworthy that even following a recruitment maneuver, subsequent consequential de-recruitment was reported only when PEEP was < 10 cm H2O.25


If one adheres to a “least PEEP” philosophy, the evidence suggests that for most patients with ARDS, targeting PEEP between 10 and 16 cm H2O probably is sufficient in all but the most severe presentations. (In our institution, we estimate this to occur in no more than 15-20% of cases). In those relatively infrequent occurrences, PEEP levels > 20 cm H2O using a fixed driving pressure of 15-20 cm H2O should be attempted to stabilize FRC and gas exchange.

This can be justified based on several observations gleaned over the past half century. First, the very early exudative phase of ARDS (initial 48 hours) is characterized by congestive atelectasis and peripheral airspace obstruction from pulmonary edema that has not yet solidified. During this period, the lungs are more amenable to recruitment and displacement of pulmonary edema out of the airspaces. Second, these cases are typified by sustained, pronounced, compressive forces emanating from both abnormal chest wall compliance and superimposed hydrostatic pressure (from edematous overlying lung tissue) favoring lung collapse. In these situations, some combination of high or super PEEP, prone positioning, and judicious use of alveolar recruitment maneuvers is indicated. This should be part of the clinician’s first-line armamentarium in treating intractable hypoxemia in unusually severe presentations of ARDS.


  1. Ashbaugh DG, Bigelow DB, Petty TL, Levine BE. Acute respiratory distress syndrome. Lancet 1967;2:319-323.
  2. Gattinoni L, Carlesso E, Cressoni M. Selecting the ‘right’ positive end-expiratory pressure level. Curr Opin Crit Care 2015;21:50-57.
  3. Kallet RH. Should PEEP titration be based on chest mechanics in patients with ARDS? Respir Care 2016;61:876-890.
  4. Falke KJ, Pontoppidan H, Kumar A, et al. Ventilation with end-expiratory pressure in acute lung disease. J Clin Invest 1972;51:2315-2323.
  5. Suter PM, Fairley HB, Isenberg MD. Optimum end-expiratory airway pressure in patients with acute pulmonary failure. N Engl J Med 1975;292:284-289.
  6. Katz JA, Ozanne GM, Zinn SE, Fairley HB. Time course and mechanism of lung-volume increases with PEEP in acute pulmonary failure. Anesthesiology 1981;54:9-16.
  7. Hubmayr RD, Kallet RH. Understanding pulmonary stress-strain relationships in severe ARDS and its implications for designing a safer approach to setting the ventilator. Respir Care 2018;63:220-227.
  8. Kallet RH, Matthay MA. Hyperoxic acute lung injury. Respir Care 2013;58:123-140.
  9. Borges JB, Okamato VN, Matos GFJ, et al. Reversibility of lung collapse and hypoxemia in early acute respiratory distress syndrome. Am J Respir Crit Care Med 2006;174:268-278.
  10. Mushin WW, Rendell-Baker L, Thompson PW, Mapledon WW. Chapter 3 in: Automatic Ventilation of the Lung. 3rd ed. London: Blackwell Science Ltd.; 1969: 62-79.
  11. Suter PM, Fairley HB, Isenberg MD. Effect of tidal volume and positive end-expiratory pressure on compliance during mechanical ventilation. Chest 1978;73:158-162.
  12. Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as a compared with traditional tidal volumes for acute respiratory distress syndrome. N Engl J Med 2000;342:1301-1308.
  13. Albert RK. Least PEEP: Primum non nocere. Chest 1985;87:2-4.
  14. Kirby RR, Downs JB, Civetta JM, et al. High level positive end-expiratory pressure (PEEP) in acute respiratory insufficiency. Chest 1975;67:156-163.
  15. Lachmann B. Open up the lung and keep the lung open. Intensive Care Med 1992;18:319-321.
  16. Brower RG, Lanken PN, MacIntyre N, et al. Higher versus lower positive end-expiratory pressures in patients with the acute respiratory distress syndrome. N Engl J Med 2004;351:327-336.
  17. Meade MO, Cook DJ, Guyatt GH, et al., for the Lung Open Ventilation Study Investigators. Ventilation strategy using low tidal volumes, recruitment maneuvers and high positive end-expiratory pressure for acute lung injury and acute respiratory distress syndrome: A randomized controlled trial. JAMA 2008;299:637-645.
  18. Mercat A, Richard J-CM, Vielle B, et al., for the Express Study Group. Positive end-expiratory pressure setting in adults with acute lung injury and acute respiratory distress syndrome: A randomized controlled trial. JAMA 2008;299:646-655.
  19. Writing Group for the Alveolar Recruitment for Acute Respiratory Distress Syndrome Trial (ART) Investigators. Effect of lung recruitment and titrated positive end-expiratory pressure (PEEP) vs. low PEEP on mortality in patients with acute respiratory distress syndrome. JAMA 2017;318:1335-1345.
  20. Talmor D, Sarge T, Malhotra A, et al. Mechanical ventilation guided by esophageal pressure in acute lung injury. N Engl J Med 2008;359:2095-2104.
  21. Jardin F, Desfond P, Bazin M, et al. Controlled ventilation with best positive end-expiratory pressure (PEEP) and high level PEEP in acute respiratory failure (ARF). Intensive Care Med 1981;7:171-176.
  22. Lemaire F, Simoneau G, Harf A, et al. Static pulmonary pressure-volume (P-V) curve, positive end-expiratory pressure (PEEP) ventilation and gas exchange in acute respiratory failure (ARF). Am Rev Respir Dis 1979;119:328 (Suppl).
  23. Kallet RH, Branson RD. Do the NIH ARDS clinical trials network PEEP/FiO2 tables provide the best evidence-based guide to balancing PEEP and FiO2 settings in adults? Respir Care 2007;52:461-477.
  24. Gattinoni L, D’Andrea L, Pelosi P, et al. Regional effects and mechanism of positive end-expiratory pressure in early adult respiratory distress syndrome. JAMA 1993;269:2122-2127.
  25. Crotti S, Masheroni D, Caironi P, et al. Recruitment and derecruitment during acute respiratory failure. Am J Respir Crit Care Med 2001;164:131-140.