现在的位置: 首页研究点评, 进展交流>正文
[JAMA发表述评]: 寻找非ARDS患者的适当PEEP:高,低或适中?
2021年01月18日 研究点评, 进展交流 暂无评论

Editorial December 9, 2020

Searching for the Optimal PEEP in Patients Without ARDS: High, Low, or in Between?

Sarina K. Sahetya, Ewan C. Goligher, Arthur S. Slutsky

JAMA. Published online December 9, 2020. doi:10.1001/jama.2020.23067

Positive pressure ventilation is lifesaving for patients with acute respiratory failure and has been the cornerstone of care for critically ill patients since the polio epidemic in the 1950s. Increasing end-expiratory pressure in ventilated patients, termed positive end-expiratory pressure (PEEP), was noted to be beneficial for acute pulmonary edema as early as the 1930s by Barach et al.1 However, it was not widely used until 1967, when Ashbaugh and colleagues2 “discovered” acute respiratory distress syndrome (ARDS) and demonstrated that PEEP improved arterial oxygenation in some patients with ARDS. For decades, the major goal of PEEP was to improve oxygenation or oxygen delivery,3 but over the past few decades, this goal has shifted to minimizing ventilator-induced lung injury, an approach that includes limiting tidal volumes and inspiratory pressures while providing sufficient PEEP to minimize lung collapse.4

From a physiological perspective, PEEP may be beneficial by maintaining lung units (alveoli and small airways) open that would otherwise collapse at end expiration, thus improving gas exchange. By keeping additional lung units open, PEEP can reduce injurious shear forces due to cyclic opening and closing of these units during ventilation, and allow distribution of tidal volumes over a larger and more homogeneously inflated lung surface, thereby reducing dynamic stress and strain. During spontaneous breathing, maintaining lung recruitment with PEEP may also decrease respiratory effort and eccentric diaphragm loading.

However, higher PEEP levels can have detrimental consequences by leading to increased inspiratory pressures and overdistention of the lung along with increased lung stress and strain, and therefore, more ventilator-induced lung injury. Higher intrathoracic pressures with PEEP can also impair hemodynamics by reducing venous return and increasing pulmonary vascular resistance. The net benefit or harm from PEEP therefore depends on this balance of alveolar recruitment to overdistension and should be particularly beneficial in disease states with substantial alveolar collapse, such as ARDS.

These trade-offs in using PEEP have been studied for decades in patients with ARDS, with 5 large randomized clinical trials of higher vs lower PEEP providing somewhat variable findings.5-9 A meta-analysis demonstrated that patients with moderate or severe ARDS—but not mild ARDS—benefited from a higher PEEP ventilation strategy.10

Much less attention has been paid to the use of PEEP in patients without ARDS and who have relatively healthy lungs. These studies have largely focused on patients at increased risk of atelectasis, eg, patients receiving anesthesia for operative procedures. The optimum PEEP level in patients without ARDS would be expected to be lower and the risk-benefit ratio higher because they have relatively less lung collapse than patients with ARDS and require less pressure to open the collapsed lung. In 2 randomized trials evaluating patients undergoing open abdominal surgery (n = 900) and obese patients undergoing surgery (n = 1976), there were no differences in postoperative pulmonary complications between an intraoperative ventilation strategy using ultra-low PEEP (<5 cm H2O) and use of a PEEP of 12 cm H2O plus recruitment maneuvers, which included incremental increases in tidal volume and/or PEEP that were repeated intermittently.11,12 In nonsurgical patients with a risk of ARDS, Pepe and colleagues13 also found no difference in subsequent development of ARDS using a PEEP of 0 cm H2O vs a PEEP of 8 cm H2O.

In this issue of JAMA, the RELAx Collaborative Group add to this literature by reporting results of a large, randomized clinical trial of higher vs lower PEEP in 980 ventilated patients.14 The study, conducted in 8 Dutch intensive care units, enrolled patients without ARDS who were expected to be intubated for more than 24 hours. Participants were randomized to receive either a lower PEEP strategy, in which PEEP was titrated close to 0 with a maximum of 5 cm H2O (n = 476), or a higher PEEP strategy of 8 cm H2O (n = 493). Fraction of inspired oxygen was titrated to a maximum of 0.6 before PEEP could be increased to maintain adequate oxygenation. Most patients enrolled were nonsurgical patients (79%); the 2 main reasons for intubation were respiratory failure (30%) and cardiac arrest (27%).

The primary outcome was the number of ventilator-free days at day 28. The investigators used a noninferiority trial design, with the noninferiority margin set at −10%, roughly equivalent to 1.6 ventilator-free days or 12 hours of ventilation, on the reasonable assumption that this difference was not clinically meaningful. They found that the lower PEEP strategy met criteria for noninferiority with a median of 18 ventilator-free days compared with 17 days in the higher PEEP strategy (mean ratio, 1.04; 95% CI, 0.95-∞; P = .007 for noninferiority). A superiority analysis also suggested no statistically significant difference between groups (P = .22).

What are the clinical implications of this trial? Although this investigation is a large, randomized, multicenter clinical trial and was rigorously conducted, there are a few challenges in interpreting the results. First, as the authors acknowledge, neither strategy they chose represents standard care. The investigators chose 8 cm H2O, rather than 5 cm H2O, as their control strategy based on evidence that “usual care” PEEP has drifted higher over time.15 However, based on other large observational studies, it appears that clinicians largely use PEEP levels less than 8 cm H2O for this patient population.16

Second, heterogeneity of treatment effect is a widely documented challenge to interpreting clinical trials of lower vs higher PEEP in patients with ARDS.17,18 This challenge may be even more pronounced in this trial of diverse non-ARDS patients. Although there were no statistically significant differences in outcomes by the evaluated subgroups, there was suggestion of possible differences in treatment effect according to the reason for intubation (P = .08 for interaction). Thus, it is possible that the lower PEEP strategy may not be noninferior in all patient populations, especially those intubated for respiratory failure.

Third, the trial was based on a noninferiority design. Noninferiority designs are usually used to evaluate new interventions that have some compelling advantage in terms of lower toxicity, cost, or effort (such as mode or frequency of administration).19 The novel intervention is typically compared with an “active control” or known effective treatment because there would be little benefit in demonstrating that the new intervention is not inferior to an unproven therapy. For example, in a novel drug trial, a lower dose may have less toxicity and also lower efficacy, so demonstrating noninferiority may be clinically relevant. In the intensive care unit, lower PEEP is achieved by turning a knob on the ventilator and is not less expensive or less logistically challenging than providing higher PEEP. Moreover, lower PEEP may have its own adverse effects (eg, atelectasis, hypoxemia) and may have less, equivalent, or greater efficacy than a higher PEEP strategy. In this study, the rates of severe hypoxemia (20.6% vs 17.6%) and the need for rescue therapy (19.7% vs 14.6%) were numerically higher (but not statistically significantly different in adjusted analysis) in the lower PEEP group, suggesting the possibility that lower PEEP may have been inferior for some patients. For an average patient without ARDS, however, the results suggest that clinicians should feel as comfortable choosing a lower PEEP as they do choosing a higher PEEP.

Ultimately, clinicians will need to decide what the results of this study mean for the care of patients receiving mechanical ventilation. A PEEP of 0 to 5 cm H2O is noninferior to a PEEP of 8 cm H2O with respect to ventilator-free days. However, given the concern about possible increased rates of hypoxemia and need for rescue strategies in the lower PEEP group, an intermediate option of 5 to 8 cm H2O that is consistent with the current PEEP management for many non-ARDS patients is likely reasonable and may be safer than a very low PEEP strategy.15

Future trials of PEEP should be designed using strategies that individualize PEEP according to a specific patient’s physiology rather than focusing simply on dose. For example, comparing high and low insulin doses is less informative than comparing tight vs liberal glucose goals in patients with diabetes. In the context of PEEP, the potential benefit of PEEP is almost certainly proportional to the degree of recruitable atelectasis. However, no trial to date has attempted to assess PEEP responsiveness in specific patients prior to randomization. Thus, it is possible, and perhaps even likely, that the beneficial effect of higher PEEP for some patients with recruitable lung tissue is nullified by the detrimental effects of overdistension and hemodynamic compromise in others. These trials should be conducted using an enriched population of patients with the highest probability of benefit rather than simply using high vs low PEEP applied uniformly to all patients who require mechanical ventilation.

给我留言

您必须 [ 登录 ] 才能发表留言!

×
腾讯微博