INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In ventilated patients with acute lung injury (ALI), positive end-expiratory pressure (PEEP) may improve pulmonary gas exchange. However, it may also decrease cardiac output and thus offset the expected benefits in terms of oxygen delivery. The PEEP-induced decrease in cardiac output is assumed to be mainly due to a decrease in systemic venous return secondary to the increased pleural pressure (1). Impairment of right ventricular (RV) ejection related to increased transpulmonary pressure (i.e., alveolar minus pleural pressure) could also play a role in some patients (4, 5). The adverse hemodynamic effects of PEEP are not easily predictable in clinical practice, although they have been shown to be more likely to occur in patients with low left ventricular (LV) filling pressure (6).

Mechanical ventilation induces cyclic changes in LV stroke volume (SV) characterized by a lower LVSV during expiration than during insufflation (9). This respiratory pattern is mainly explained by the expiratory decrease in LV filling that followed after a delay (caused by the long pulmonary transit time of blood) the decrease in RVSV occurring during insufflation (10). The inspiratory decrease in RVSV has been shown to result essentially from a decrease in RV filling caused by the effects of increased pleural pressure on systemic venous return (9) and from transient impairment of RV ejection related to increased transpulmonary pressure on pulmonary circulation (13, 14).

Interestingly, the decrease in mean cardiac output induced by PEEP and the decrease in RVSV induced by mechanical insufflation share the same mechanisms, i.e., the negative effects of increased pleural pressure on RV filling and of increased transpulmonary pressure on RV ejection. Thus, it is reasonable to expect that the magnitude of the expiratory decrease in LVSV would correlate with the PEEP-induced decrease in mean cardiac output.

Finally, the negative effects of increased pleural pressure on RV filling should be more pronounced in patients with low cardiac preload (15, 16). Thus, the beneficial effect of fluid loading on cardiac output might be expected to correlate with the magnitude of the inspiratory decrease in RVSV and hence of the expiratory decrease in LVSV before fluid loading.

Aortic pulse pressure is directly proportional to LVSV and inversely related to aortic capacitance (17). Respiratory changes in peripheral pulse pressure (PP) during mechanical ventilation have been shown to closely reflect the variations in LVSV during the respiratory cycle (10). Thus, the aim of our study was to examine the relationships between PP and the hemodynamic effects of PEEP and fluid loading in ventilated patients with ALI. We hypothesized that the higher the PP on ZEEP, the higher the PEEP-induced decrease in cardiac output. In patients who received fluid while on PEEP, we also hypothesized that the higher the PP before fluid loading, the higher the fluid-loading-induced increase in cardiac output.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The protocol was approved by the institutional review board for human subjects (Comité Consultatif de Protection des Personnes dans la Recherche Biomédicale, Cochin Hospital), and written informed consent was obtained from all the patients' next of kin.

Patients

We studied 14 mechanically ventilated patients in whom ALI was diagnosed. This group consisted of 10 men and four women 37 to 83 yr of age (mean age, 58 ± 16 yr).

Inclusion criteria were as follows: (1) ALI defined by the combination of recent bilateral pulmonary infiltrates on chest radiograph, a Pa_O2/FI_O2 ratio < 300 mm Hg, and a pulmonary artery occlusion pressure (Ppao) below 18 mm Hg; (2) all patients had to be instrumented with indwelling arterial (radial or femoral) and pulmonary artery catheters; (3) and all patients had to be hemodynamically stable, as defined by a variation in heart rate, blood pressure, and CI of less than 10% over the 15-min period before starting the protocol. Patients were excluded if they had arrhythmias or any contraindication to the use of PEEP.

Hemodynamic Measurements

Patients were studied while supine, and zero pressure was measured at the midaxillary line. Right atrial pressure (PRA) and Ppao were recorded throughout the respiratory cycle and measured at end-expiration. Cardiac output was calculated as the mean of five measurements obtained by injecting 10 ml of dextrose solution randomly during the respiratory cycle. The CI was calculated as the ratio of cardiac output to body surface area.

Arterial Pressure Variations

We used the analog output from the monitor (H-P Monitor M1092A; Hewlett-Packard, Les Ullis, France) via an A-T-D converter to record the arterial pressure and airway pressure curves over at least 10 breaths simultaneously onto a computer (Toshiba 3200 SX). Recording was performed at a sampling rate of 500 Hz using customized acquisition software. Pulse pressure (PP) was calculated on a beat-to-beat basis as the difference between systolic and diastolic arterial pressure. Maximal PP (PPmax) and minimal PP (PPmin) values were determined over a single respiratory cycle. To assess the respiratory changes in PP, the percent change in PP was calculated as:

An example of our data and their analysis is shown in Figure 1.

View larger version (17K): [in this window] [in a new window]   Figure 1.   Simultaneous recording of systemic arterial and airway pressure curves in one illustrative patient with large pulse pressure (PP) variations. Pulse pressure was calculated on a beat-to-beat basis as the difference between systolic and diastolic arterial pressure. Maximal (PPmax) and minimal (PPmin) values for PP were determined over a single respiratory cycle. Respiratory pulse pressure variations (PP) were calculated as: PP (%) = 100 × {(PPmax  PPmin)/([PPmax + PPmin]/2)}.

Respiratory Measurements

Airway pressures were measured by using a pressure transducer (Uniflow 43-600; Baxter Edwards Crit Care, Irvine, CA) connected close to the proximal end of the endotracheal tube. Plateau airway pressure (Pplat) was measured after an end-inspiratory (2 s) occlusion. Tidal volume (VT) was measured by means of the ventilator transducer. The static compliance of the respiratory system (Cst,rs) was calculated as follows: Cst,rs = VT/(Pplat  PEEP).

Study Protocol

All patients were sedated and mechanically ventilated in a volume-controlled mode with an I/E ratio of one-half to one-third. Six patients were therapeutically paralyzed according to the attending physician. In three of the eight remaining patients, spontaneous breathing activity was detected by visual inspection of the airway pressure curve. To ensure that PP reflected only the effects of positive pressure ventilation, these three patients were temporarily paralyzed. Measurements were performed in duplicate, first during 0 cm H₂O PEEP (ZEEP) and then 15 min after the addition of 10 cm H₂O PEEP (PEEP). In patients in whom PEEP induced a decrease in CI of at least 10%, fluid loading using 500 ml Hetastarch was performed over 30 min and a third set of hemodynamic measurements was then obtained. Except for PEEP, ventilatory settings and dosages of inotropic and vasopressive drugs were held constant.

Statistical Analysis

Results were expressed as means ± standard deviation. The effects of PEEP and fluid loading were assessed using Wilcoxon's nonparametric rank sum test (18). Correlations were tested using Spearman's rank test. A p value less than 0.05 was considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The main characteristics of the 14 patients studied are listed in Table 1. Our patients had no history of heart failure, and CI during ZEEP ranged from 2.9 to 7.0 L/min/m². All patients exhibited maximal PP during insufflation and minimal PP during the expiratory period. The effects of PEEP on the hemodynamic parameters are presented in Table 2.

On ZEEP, PP correlated both with PRA (r = 0.62, p < 0.05) and with Ppao (r = 0.64, p < 0.05). However, PP on ZEEP did not correlate with VT and Crs,st.

PEEP induced a decrease in CI from 4.2 ± 1.1 to 3.8 ± 1.3 L/min/m² (p < 0.01) and an increase in PP from 9 ± 7 to 16 ± 13% (p < 0.01). The PEEP-induced changes in CI correlated both with PP on ZEEP (r = 0.91, p < 0.001) and with the PEEP-induced changes in PP (r = 0.79, p < 0.001) (Figure 2). The PEEP-induced changes in CI also correlated with Ppao on ZEEP (r = 0.75, p < 0.01) but were not significantly correlated with PRA on ZEEP (r = 0.48, p = 0.08).

View larger version (11K): [in this window] [in a new window]   Figure 2.   (Upper panel ) Correlation between respiratory pulse pressure variations (PP) on ZEEP and the PEEP-induced changes in cardiac index (CI). (Lower panel ) Correlation between PEEP- induced changes in PP (PP on PEEP  PP on ZEEP) and the PEEP-induced changes in CI.

Six patients demonstrated a decrease in CI > 10% with the application of PEEP. In these patients, fluid loading increased CI from 3.5 ± 1.1 to 4.2 ± 0.9 L/min/m² (p < 0.05) and decreased PP from 27 ± 13 to 14 ± 9% (p < 0.05). The fluid-loading-induced changes in CI correlated with PP on PEEP before volume expansion (r = 0.97, p < 0.01) and with the fluid-loading-induced changes in PP (r = 0.85, p < 0.05) (Figure 3).

View larger version (11K): [in this window] [in a new window]   Figure 3.   (Upper panel ) Correlation between respiratory pulse pressure variations (PP) on PEEP before fluid loading (FL) and the fluid loading-induced changes in cardiac index (CI). (Lower panel ) Correlation between fluid loading-induced changes in PP (PP after FL  PP before FL) and the fluid loading-induced changes in CI.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our results demonstrate strong relationships between PP and the effects of both PEEP and fluid loading on cardiac output in ventilated patients with ALI.

All of our patients exhibited a maximal PP during mechanical insufflation and a minimal PP at expiration. These findings are consistent with the respiratory pattern of arterial pressure previously described in animal and clinical studies during positive pressure ventilation (9, 19). The respiratory changes in PP have been shown related to the cyclic changes in LVSV that followed after a delay the respiratory changes in RVSV (10). At insufflation, RVSV is minimal because of the negative effects both of increased pleural pressure on RV filling (9) and of increased transpulmonary pressure on RV ejection (13, 14). In conventional ventilatory conditions, this should result in a minimal LVSV during expiration because of the phase lag between RV output and LV filling caused by the long blood pulmonary transit time (9, 10, 21). Other mechanisms might also contribute to the increase in LVSV at insufflation, particularly in patients with congestive heart failure: (1) a further LV filling caused by squeezing of blood out of alveolar vessels (20, 27), and (2) a decrease in LV afterload caused by the increased pleural pressure (11, 28).

In fact, the main mechanisms that induce the inspiratory decrease in RVSV and hence the expiratory decrease in LVSV are identical to those whereby PEEP decreases mean cardiac output. Accordingly, we found a strong correlation between PP on ZEEP and the PEEP-induced decrease in CI. This finding suggests that PP may be useful in predicting the hemodynamic effects of PEEP.

The expiratory decrease in LVSV caused by reduced LV preload should be greater when the left ventricle operates on the steep rather than on the flat portion of the Frank-Starling curve (15, 16). Similarly, the inspiratory decrease in RVSV would be greater in the case of low RV filling (15, 16). These combined phenomena have been proposed to explain why respiratory changes in arterial pressure are either increased by hemorrhage (22, 23) or decreased by fluid loading (23, 25) and why they correlate with LV preload indices such as Ppao (24) and LV end-diastolic area (25). We also found that PP correlated with PRA and Ppao. Furthermore, in the six patients who received fluid, the increase in CI correlated both with PP before fluid loading and with the fluid-loading-induced decrease in PP. These findings suggest that PP may be useful for monitoring the hemodynamic effects of fluid loading.

No correlation was found between PP and tidal volume. This result could be due to the small range of VT and to the fact that, in contrast to others studies (19, 23), we did not modify VT throughout the study.

When our patients were transfered from ZEEP to PEEP, the changes in PP strongly correlated with the changes in CI. These results were in accordance with those of Pizov and colleagues (26) who found that systolic pressure variations in dogs increased mostly when cardiac output decreased with PEEP. In preload-sensitive subjects, it may be assumed that the further increase in pleural pressure with PEEP would have produced a greater decrease in both expiratory LVSV and mean cardiac output. However, because our study was not designed to elucidate why PP increased with PEEP, we cannot exclude the possibility that mechanisms affecting RV afterload may also have occurred: an additional increase in RV afterload during insufflation on PEEP, related to the extension of West's Zone 1 or 2 (13) cannot be excluded. Conversely, PEEP-induced improvement in functional residual capacity and/or a decrease in hypoxic pulmonary vasoconstriction might have resulted in a lower RV afterload during insufflation on PEEP than on ZEEP.

It must be underlined that arrhythmias and spontaneous breathing activity may result in misleading interpretation of PP. Finally, since our study concerned patients with ALI, the results cannot be extrapolated to patients with chronic respiratory disease or congestive heart failure.

In summary, our findings suggest that (1) PP could be used at the bedside to predict adverse hemodynamic effects of PEEP, (2) changes in PP from ZEEP to PEEP could be used to assess changes in CI that occur when PEEP is applied, (3) in patients with ALI ventilated with PEEP, PP and its changes induced by fluid may be helpful in predicting and assessing the effects of fluid loading on hemodynamics. Because the potential risk of using pulmonary artery catheters is currently a subject of debate (29), the use of PP to monitor hemodynamics in ventilated patients with ALI may be an attractive alternative approach.