© 2003 American Thoracic Society
Auto-peep with low tidal volumeTo the Editor:We read with interest the paper by de Durante and colleagues (1) implicating high levels of intrinsic positive end-expiratory pressure (PEEPi) as a possible explanation for the success of the ARDSnet low tidal volume strategy (2). It is satisfying to learn of a possible link between the results of that clinical trial and extensive laboratory experience that documents the contribution of elevated PEEP to a lung protective strategy when high inflation pressures are used. However, some of the data in this report suggest alternative and quite different explanations for the authors' observations. Modest auto-PEEP sometimes develops in acutely injured lungs, especially at low lung volumes. This is an understandable consequence of relatively high quantities of air flowing through a reduced number of airways and the compressive forces of heavy lungs and mediastinal weight on dependent bronchi. Importantly, these low levels of genuine auto-PEEP (and evidence for flow limitation) are predictably obliterated by the application of < 10 cm H2O of extrinsic PEEP (3)—the value used in the de Durante study. Moreover, assuming unchanging respiratory system and expiratory circuit mechanics, auto-PEEP should not rise markedly with increasing frequency, so long as expiratory time fraction (PE/TTOT), and minute ventilation (the primary determinants of expiratory flow) remain the same—as they did in this clinical experiment (4). Figure 1B of the paper in question (1) demonstrates an extraordinarily high circuit pressure on the truncated tracings immediately before the first inflation cycle with the "lower VT" strategy. This reflects the well known high resistance of the Servo 900C scissor valve (5). Expiratory retardation and inadvertent PEEP originating in this equipment are inevitable at high operating frequencies. More recently engineered ventilators offer dramatically improved valve systems. Fortunately, few 900C ventilators were used in the ARDSnet trial. A final troublesome point: the small difference (8.2 cm H2O) between plateau pressure and total PEEP at low VT implies a thoracic compliance that was uncharacteristically high for patients with acute respiratory distress syndrome (ARDS). Thus, we do not think the data presented in this paper provide strong evidence for substantially higher total-PEEP in the lower tidal volume group of the ARDSnet trial. It is notable also that PaO2/FIO2 ratios and respiratory system compliances were lower in the lower tidal volume group of the ARDSnet trial. These changes suggest that the lower tidal volume strategy was associated with more atelectasis, presumably from the smaller tidal swings in lung volume and transpulmonary pressure. Resolution of these questions seems important and timely, given the reported failure of higher PEEP to improve ARDS mortality in the now completed ARDSnet trial testing that question (6).
a Regions Hospital St. Paul, Minnesota REFERENCES
From the Authors:We thank Drs. Marini and Brower for their thoughtful comments.Auto–positive end expiratory pressure (PEEP) can be caused by (1) reduced elastic recoil, (2) increased resistance, (3) excessive tidal volume, (4) high respiratory rate and/or short duty cycle, (5) expiratory flow limitation (1). Auto-PEEP due to any of these mechanisms will lead to an increase in end-expiratory lung volume with the potential to minimize ventilator-induced lung injury. Although minute ventilation and expiratory time fraction are likely to be the prime determinants of auto-PEEP, they are not the only determinants. The mathematical model Marini and Brower cite (2) is based on a number of assumptions, including linearity and a single compartment model, both of which are, at best, partially valid in patients with acute respiratory distress syndrome (ARDS). Marini and Brower suggest that development of auto-PEEP may be related to a specific ventilator. However, Veillard-Baron (3) found that patients ventilated with a tidal volume of 400–500 ml and a respiratory rate of 30 breaths per minute generated 6.4 ± 2.7 cm H2O of auto-PEEP; all patients were ventilated with a Puritan-Bennett 7200. Low chest wall compliance may generate flow limitation in ARDS because of the compressive effect of high intrathoracic pressure (1). We agree that our patients had a normal thoracic compliance, making expiratory flow limitation as the mechanism responsible for auto-PEEP unlikely, but as stated above, increased end-expiratory lung volume would still ensue. Marini and Brower point out that oxygenation and compliance were lower (on average) in the low tidal volume group in the original ARDS network trial (4). This is not surprising given our results, since we found that significant auto-PEEP only develops at very high respiratory rates (5), and it is only in this minority of patients that there might have been increased auto-PEEP with improved oxygenation and compliance; examination of the average values is therefore not as instructive as looking at individual values, since oxygenation and compliance would likely have been lower in the majority of patients in the low tidal volume group at lower respiratory rates in whom auto-PEEP would not have occurred. A recently completed trial assessed whether a protective strategy that included levels of PEEP higher than in the original ARDS network trial (4) may improve outcome: failure of high PEEP to reduce mortality was reported (6). Nevertheless, significant differences in the baseline patient characteristics were described; after correction for these baseline imbalances, the higher PEEP group had a relative decrease in mortality of about 10% (6). This decrease was not significant, but the study was not powered to detect a difference of this magnitude (6).
a Università di Torino Torino, Italy REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
