Dynamic Intrinsic PEEP (PEEP_i,dyn)
Is It Worth Saving?

Magdy Younes

Department of Internal Medicine, University of Manitoba, Manitoba, Winnipeg, Canada

	ARTICLE

TOP ARTICLE REFERENCES

Dynamic hyperinflation (DH) refers to failure of lung volume to return to passive FRC before the onset of inspiration. The presence of DH during assisted ventilation results in increased work of breathing, poor patient-ventilator interactions (delayed triggering and ineffective efforts), and errors in estimating respiratory mechanics and patient respiratory rate (1).

Since the early 1990s, the reduction in esophageal pressure (Pes) between the onset of inspiratory effort and onset of inspiratory flow (PEEP_i,dyn) has been used to estimate DH in patients with respiratory efforts (5). More recently, the validity of PEEP_i,dyn has been questioned on two grounds:

1. There is evidence that PEEP_i,dyn underestimates the true magnitude of DH (5, 8, 9), at times by as much as 90% (8). This is largely related to nonhomogeneous mechanical properties within the lungs. The pressure at which flow becomes inspiratory is determined primarily by lung units with the lowest alveolar pressure (Palv). The latter is necessarily lower than average Palv.

2. It is now well established that many patients generate expiratory pressure (Pexp) near end expiration (10), which causes PEEP_i,dyn to overestimate DH (10). Two groups of investigators (11, 12) proposed three approaches to correct PEEP_i,dyn for the contribution of Pexp. There is uncertainty about whether any of these approaches is accurate and, if so, which one. In this issue of the Journal Zakynthinos and coworkers (13) compare these three correction methods with the correction obtained using the Campbell diagram. Before discussing the results of this study, it is useful to consider what information is needed to accurately offset the effect of Pexp and why the Campbell diagram is, theoretically, the ideal arbitrator.

In a passive system, Palv reflects elastic recoil (Pel), which is a function of volume above passive FRC. At a given volume, however, Palv can be increased or decreased by expiratory or inspiratory muscle action, respectively. At end expiration (t₀), Palv is given by Eq. (1):

(1)

It is Pel at t₀ that we need to determine in order to estimate DH. With onset of inspiratory effort, Pinsp increases and Pexp decreases. As a result, Palv decreases. When Palv crosses external PEEP, flow crosses zero and becomes inspiratory. The change in Palv, as reflected in Pes, between t₀ and zero flow crossing, therefore, is a measure of how much Palv exceeded PEEP at t₀ (i.e., intrinsic PEEP). Having determined Palv at t₀, the missing information is the net muscle pressure (Pexp  Pinsp) at t₀. What combination of Pexp and Pinsp was used to reduce Palv in the interval between t₀ and flow crossing has no bearing whatsoever on what net Pmus (in absolute terms) was at t₀.

The Campbell diagram is, theoretically, ideal for determining net Pmus at any point in the respiratory cycle. The relation between intrathoracic pressure (Pes) and lung volume in the passive state is defined. Any deviation of actual Pes during active breathing from the passive value at the same volume is a measure of net Pmus.

In all three approaches proposed to correct for Pexp, gastric pressure (Pga) at t₀ is measured. Because absolute Pga is meaningless, a reference value must be subtracted to arrive at Pmus. The three approaches use different values of Pga as reference. The uncertainties surrounding these approaches have to do with whether Pga at t₀ reflects Pes at t₀, and which, if any, of the reference values reflects what Pes would be in the passive state under the same conditions as at t₀. The first uncertainty would not be of concern if diaphragm tension at t₀ is near zero. Although this is reasonable in most cases, tonic diaphragm activity has been reported during experimental hyperinflation (14, 15). Furthermore, development of passive tension in the diaphragm, as it is pushed up by abdominal muscles, cannot be excluded. The second uncertainty is more problematic. Pga is a complex variable. It is affected by total volume, by chest wall configuration, and by the action of either inspiratory or expiratory muscles. All three reference points differ from the condition at t₀ in overall lung volume and chest wall configuration and an indeterminate amount of inspiratory pressure may be present at any of them. Accordingly, none of these approaches is theoretically defensible. It is possible, however, that one of them may, because of specific features of responses in mechanically ventilated patients, be close to the truth. This is what Zakynthinos and coworkers (13) were testing, utilizing the Campbell diagram as the "gold" standard. They showed excellent agreement between the results of the Campbell diagram and the results of one of the Lessard approaches (Pga, total decay). The other two approaches did not correlate as well, and differences were sometimes large.

I have two concerns that temper my enthusiasm for the authors' conclusion. First, the Campbell diagram is, in practice, not a precise instrument because passive and active data are collected under different conditions. With respect to the current study, one cannot be absolutely sure that the dynamic loop was not positioned somewhat higher or lower than it really was, relative to the volume axis. Furthermore, passive chest wall pressure may have changed because of differences in thoracic blood volume (due to differences in intrathoracic pressure) or due to chest wall distortion. Distortion is evident in most reported examples (12, 13, 15). Changes in the reference value of Pes, as a result of the catheter sliding up or down the esophagus or because of a change in the compressive effect of the heart (due to a change in body position or heart size), cause errors in the horizontal placement of the loop and these translate, on a 1:1 basis, into errors in estimating Pmus and DH. These uncertainties make it imprudent to expect estimated DH to have a precision of less than ± 2-3 cm H₂O, and this is likely a charitable assessment. Such uncertainty may be tolerable under most circumstances where the diagram is used. In this case, however, the diagram is called on to arbitrate between results that differ from each other by only a few centimeters of water. Small differences between the active and passive states, particularly when systematic, may greatly favor one method.

The second concern relates to a peculiar inconsistency in the data. In Table 2 of Zakynthinos and coworkers (13), estimated DH using the Appendini approach is given as 8.5 ± 3.1 cm H₂O. Elsewhere in the article, the authors report Pdi between t₀ and zero flow crossing as being 5.3 ± 0.6 cm H₂O. Yet, the two values are mathematically bound to be identical (both are Pes  Pga over the same interval). Thus, whether the Appendini approach yielded 8.5 ± 3.1 or 5.3 ± 0.6 cm H₂O is not clear. Considering that the Campbell diagram yielded 5.3 ± 1.9 cm H₂O, it is difficult to say whether the Appendini approach gave, on average, the best or worst results. This inconsistency must be resolved before Pga, total decay is adopted as the standard method.

In my view, the most important findings of the current study are that the differences between the three methods are rather small (2-3 cm H₂O) and that each of them largely corrects for any Pexp that may exist at t₀. Considering that measurements were made near the point of exhaustion in a weaning trial, where respiratory drive and expiratory muscle use were likely high, it is likely that the differences are even smaller under most circumstances. This is encouraging and, in a way, makes the issue of which method is better less critical. The more important issue is how much quantitative uncertainty remains after applying the Pexp correction. The results of the present study (notably Figure 3A; see Zakynthinos and colleagues [13]) suggest that it is the same as the uncertainty about Campbell diagram analysis when used under the circumstances of the current study. It is difficult to estimate how much (in centimeters of water) this uncertainty is and, thus, whether it is clinically acceptable. It is, however, worth noting that the magnitude of PEEP_i,dyn, after correction for Pexp, is modest (5.6 ± 2.8 [11], 2.6 ± 3.2 [12], 5.3 ± 1.9 [13]), so that uncertainty of only a few centimeters of water represents a large fraction of the primary signal.

Is it worth pursuing more precise ways of estimating Pexp at t₀? I believe not. This would be a daunting task in a sophisticated physiology laboratory, inconceivable in the clinical setting. Even if a method is found, we would still be left with a measurement that is liable to underestimate true DH by a highly variable and indeterminate amount (see point 1, above). It can certainly be argued that the quality and utility of the information provided by this measurement do not justify an invasive procedure. Perhaps it is time to put PEEP_i,dyn to rest, and to focus on developing noninvasive ways to detect DH and to guide therapeutic approaches to minimize its consequences. This task is, in my view, more feasible, and likely to be more productive, than rehabilitating PEEPi,dyn.

Acknowledgments: Supported by grants HL57011 and HL59823 from the National Institutes of Health.

Supported by the Medical Research Council of Canada.

	References

TOP ARTICLE REFERENCES

1. Marini JJ. Dynamic hyperinflation. In: Marini JJ, Slutsky AS, editors. Lung biology in health and disease, vol 118: Physiological basis of ventilatory support. New York: Marcel Dekker; 1998. p. 453-490.

2. Parthasarathy S, Jubran A, Tobin MJ. Cycling of inspiratory and expiratory muscle groups with the ventilator in airflow limitation. Am J Respir Crit Care Med 1998; 158: 1471-1478 [Abstract/Free Full Text].

3. Giannouli E, Webster K, Roberts D, Younes M. Responses of ventilator dependent patients to different levels of pressure support and proportional assist. Am J Respir Crit Care Med 1999; 159: 1716-1725 [Abstract/Free Full Text].

4. Rossi A, Gottfried SB, Zocchi L, Higgs BD, Lennox S, Calverley PM, Begin P, Grassino A, Milic-Emili J. Measurement of static compliance of the total respiratory system in patients with acute respiratory failure during mechanical ventilation: the effect of intrinsic positive end-expiratory pressure. Am Rev Respir Dis 1985; 131: 672-677 [Medline].

5. Pertrof BJ, Legare M, Goldberg P, Milic-Emili J, Gottfried SB. Continuous positive airway pressure reduces work of breathing and dyspnea during weaning from mechanical ventilation in severe chronic obstructive pulmonary disease. Am Rev Respir Dis 1990; 141: 281-289 [Medline].

6. Dal Vecchio L, Polese G, Poggi R, Rossi A. "Intrinsic" positive end-expiratory pressure in stable patients with chronic obstructive pulmonary disease. Eur Respir J 1990;3:74-80.

7. Brochard L, Isabey D, Piquet J, Amaro P, Mancebo J, Messadi AA, Brun-Buisson C, Rauss A, Lemaire F, Harf A. Reversal of acute exacerbations of chronic obstructive lung disease by inspiratory assistance with a face mask. N Engl J Med 1990; 323: 1523-1530 [Abstract].

8. Maltais F, Reissmann H, Navalesi P, Hernandez P, Gursahaney A, Ranieri VM, Sovilj M, Gottfried SB. Comparison of static and dynamic measurements of intrinsic PEEP in mechanically ventilated patients. Am J Respir Crit Care Med 1994; 150: 1318-1324 [Abstract].

9. Hernandez P, Navalesi P, Maltais F, Gursahaney A, Gottfried SB. Comparison of static and dynamic measurements of intrinsic PEEP in anesthetized cats. J Appl Physiol 1994; 76: 2437-2442 [Abstract/Free Full Text].

10. Ninane V, Yernault JC, De Troyer A. Intrinsic PEEP in patients with chronic obstructive pulmonary disease: role of expiratory muscles. Am Rev Respir Dis 1993; 148: 1037-1042 [Medline].

11. Appendini L, Patessio A, Zanoboni S, Carone M, Gukov B, Donner CF, Rossi A. Physiologic effects of positive end-expiratory pressure and mask pressure support during exacerbations of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1994; 149: 1069-1076 [Abstract].

12. Lessard MR, Lofaso F, Brochard L. Expiratory muscle activity increases intrinsic positive end-expiratory pressure independently of dynamic hyperinflation in mechanically ventilated patients. Am J Respir Crit Care Med 1995; 151: 562-569 [Abstract].

13. Zakynthinos SG, Vassilakopoulos T, Zakynthinos E, Mavrommatis A, Roussos C. Contribution of expiratory muscle pressure to dynamic intrinsic PEEP. Am J Respir Crit Care Med 2000; 162: 1372-1375 [Abstract/Free Full Text].

14. Martin J, Powell E, Shore S, Emrich J, Engel LA. The role of the respiratory muscles in the hyperinflation of bronchial asthma. Am Rev Respir Dis 1980; 121: 441-447 [Medline].

15. Yan S, Kayser B, Tobiasz M, Sliwinski P. Comparison of static and dynamic intrinsic positive end-expiratory pressure using the Campbell diagram. Am J Respir Crit Care Med 1996; 154: 938-944 [Abstract].