Nov 3, 1994 - Effects of positive end-expiratory pressure on right ventricular function in COPD patients during acute ventilatory failure. Received: 3 November ...
Intensive Care Med (1996) 22:923-932 9 Springer-Verlag 1996
M. Dambrosio G. Cinnella N. Brienza VoM. Ranieri R. Giuliani F. Bruno T. Fiore A. Brienza
Received: 3 November 1994 Accepted: 12 March 1996
Effects of positive end-expiratory pressure on right ventricular function in COPD patients during acute ventilatory failure
Abstract Objective: To examine the effects of external positive end-expiratory pressure (PEEP) on right ventricular function in chronic obstructive pulmonary disease (COPD) patients with intrinsic P E E P (PEEPi). Design: Prospective study. Setting: General intensive care unit in a university teaching hospital. Patients: Seven mechanically ventilated flow-limited C O P D patients (PEEPi = 9.7 + 1.3 cmH20, m e a n + SD) with acute respiratory failure. Intervention: H e m o d y n a m i c and respiratory mechanic data were collected at four different levels of P E E P ( 0 - 5 - 1 0 - 1 5 c m H 2 0 ).
Measurements and results:
This study was supported in part by a grant from the Consiglio Nazionale delle Ricerche, Italy M. Dambrosio (~) 9 G. Cinnella N. Brienza 9 V.M. Ranieri 9 R. Giuliani E Bruno 9 T. Fiore 9 A. Brienza Istituto di Anestesiologia e Rianimazione, Universitg degli Studi di Bari, Policlinico, Piazza Giulio Cesare, 1-70124 Bari, Italy
H e m o d y n a m i c parameters were obtained by a Swan-Ganz catheter with a fast response thermistor. Cardiac index (CI) and end-expiratory lung volume (EELV) reductions started simultaneously when the applied P E E P was approximately 90~ of PEEPi measured on 0 c m H 2 0 (ZEEP). Changes in transmural intrathoracic pressure (PEEPi,cw) started only at a P E E P value much higher (120%) than PEEPi. The reduction in CI was related to a decrease in the right enddiastolic ventricular volume index
(RVEDVI) (r = 0.61; p < 0 . 0 0 1 ) . No correlation between CI and transmural right atrial pressure was observed. The RVEDVI was inversely correlated with PEEP-induced changes in EELV (r = - 5 5 ; p < 0.001), but no with PEEPi,cw (r = - 0 . 0 8 ; NS). The relationship between RVEDVI and right ventricular stroke work indexl considered an index of contractility, was significant in three patients, i.e., P E E P did not change contractility. In the other patients, an increase in contractility seemed to occur. Conclusions: In C O P D patients an external P E E P exceeding 90% of PEEPi causes lung hyperinflation and reduces the CI due to a preload effect. The reduction in RVEDVI seems related to changes in EELV, rather than to changes in transmural pressures, suggesting a lung/heart volume interaction in the cardiac fossa. Thus, in C O P D patients, application of an external P E E P level lower than PEEPi may affect right ventricular function. Key words Chronic obstructive pulmonary disease 9 Lung volume 9 Positive end-expiratory pressure 9 Right ventricular volumes 9 Right ventricular function
924
Introduction The effects of positive end-expiratory pressure (PEEP) on right ventricular (RV) function have been extensively investigated both in experimental and clinical studies of acute respiratory failure. The application of P E E P has also been proposed in chronic obstructive pulmonary disease (COPD) patients with intrinsic P E E P (PEEPi) in order to reduce the inspiratory workload imposed by PEEPi [1]. However, few data are available about the effects of P E E P on RV function during acute respiratory failure in C O P D patients with intrinsic PEEPi [2, 3]. In C O P D patients a condition of dynamic pulmonary hyperinflation can occur and a positive pressure known as intrinsic P E E P can be present at end-expiration. Intrinsic P E E P may be due to both a) expiratory flow limitation, related to the development of critical closure of the small airways [4, 5] and/or b) increased expiratory resistances of the small airways [6]. In C O P D patients with expiratory flow limitation, the application of external P E E P during assisted mechanical ventilation is able to counterbalance and reduce the inspiratory threshold load imposed by PEEPi, without causing further hyperinflation [1]. However, in C O P D patients a significant heterogeneity of flow limitation amongst regional lung units may occur [7]. This implies that the value of external P E E P above which changes in end-expiratory lung volume (EELV) are expected to occur (Pcrit) may be less than the PEEPi measured on P E E P equal to 0 c m H 2 0 (ZEEP). Indeed, lung hyperinflation has been found in mechanically ventilated C O P D patients at levels of external P E E P lower than PEEPi [8]. As expected, P E E P application decreases cardiac output (CO) in C O P D patients [2, 3, 8]. The decrease in CO was observed not only at levels of external P E E P equal to [2] or higher [3], but also lower, than PEEPi [8]. This reduction might be due not only to an increase in intrathoracic pressure, but also to the increase in lung volume caused by external PEEP. The PEEP-induced decrease in CO has been related to several mechanisms, the most important of which is the reduction in venous return. Such reduction has been associated with an increase in intrathoracic pressure acting on pericardial pressure and limiting diastolic filling [9], or with a direct compression of the lung volume on the cardiac fossa with subsequent increase in juxtacardiac pressure [10], or with the effect of P E E P on the abdominal venous compartment [11, 12], all mechanisms acting on the right ventricular preload. Another explanation for the reduction in CO is that the increased alveolar volume, caused by PEEP, might produce a Starling resistor effect on the pulmonary microcirculation, thus increasing the pressure load for RV ejection [13] and causing RV failure. To our knowledge, it is not known which of these mechanisms is predominant in decreasing CO in C O P D patients during P E E P application.
The aim of this study was to investigate how P E E P affects RV function in patients with severe COPD. The study was performed in mechanically ventilated patients in order to evaluate the actual effects of P E E P without the preload interference usually produced by spontaneous breathing. We hypothesize that, in C O P D patients, most of the hemodynamic alterations induced by P E E P on RV volumes are related to lung volume variations.
Patients and methods Patients Seven COPD patients with acute ventilatory failure (AVF) (age 69.4 + 4.3 mean • SD), admitted to the Intensive Care Unit (ICU) of the University Hospital in Bari, were studied. The study protocol was approved by the local Ethics Committe and each patient or next of kin gave informed consent. The diagnosis of COPD was made by physical examination and clinical tests and confirmed by patient history and previous pulmonary function tests. On admission to the ICU and before mechanical ventilation was started, all the patients were severly hypoxemic (70 mmHg). The causes of AVF were pneumonia (2), lung viral infection (3) and congestive heart failure (2). On admission to the ICU the two patients with heart failure presented pulmonary edema associated with systemic hypertension and arterial occlusion pressure (Ppao) higher than 18 mmHg. Arterial occlusion pressure was less than 15 mmHg at the beginning of the study. All patients were intubated and mechanically ventilated (Servo Ventilator 900C Siemens Elema AB, Berlin, FRG) for a mean• period of 3.1_+1.7 days (range 1-6 days).
Experimental procedures The investigation was performed with patients in the supine position after sedation (diazepam 0.2 mg/kg and fentanyl 2-3 3~/kg) and muscle paralysis (vecuronium bromide 0.1 mg/kg per h). The patients were ventilated in constant inspiratory flow mode with a tidal volume of 9-10 mg/kg, respiratory rate of 12-14 b/min, an inspiratory/expiratory ratio of 0.33. The ventilatory setting was not modified throughout the procedure, except for the level of PEEP. Positive end-expiratory pressure levels of 0, 5, 10, and 15 cmH20 were applied and maintained constant for 30 min before measurement. A specific software was used to randomize the order of the PEEP level. A physician not involved in the experimental procedure was always to provide for patient care. A 20 gauge radial arterial catheter (Arrow International Inc., Reading, PA, Model RA-04020-E) was inserted percutaneously to measure systemic arterial pressure (Pa). A pulmonary arterial Swan-Ganz catheter with a fast-reponse thermistor (Baxter-Edwards, Santa Anna, CA, Model 93A-431H 7.5 F) was inserted into the pulmonary artery from the right internal jugular vein or the left subclavian vein. Pulmonary arterial pressure (Ppa), Ppao and right atrial pressure (Pra) were measured by the Swan-Ganz catheter. The positioning of the catheter tip in the pulmonary artery was guided by the wave morphology during its introduction. The positioning in the segment of lung reflecting Zone 3 condition was achieyed using the Teboul method [14]. In order to correctly measure both CO and right ventricular ejection fraction (RVEF), the hole for injection was positioned 2 cm above the tricuspid valve [t5]. A heated pneumotachograph (Fleish no. 1, Fleish, Lausanne, Switzerland), connected to a differential pressure transducer
925
(Validyne M P 45_+2cmH20; Validyne Co., Northridge, CA), was inserted between the ?-piece of the ventilator circuit and the endotracheal tube to measure airway flow. Airway pressure (Paw) was measured near the proximal tip of the endotracheal tube (Validyne M P 45 _+100 cmH20 ). Special care was taken to avoid gas leakage in the equipment, particularly around the cuff [6]. The esophageal pressure (Peso) was measured using a double-lumen nasogastric tube 0 . 4 F r e n c h ; 127 cm; Mallinckrodt Inc., Argyle, NY) with a thin-walled vinyl balloon (10 cm long, 3.8 cm circumference) connected to a differential transducer (Validyne M P 45 + 100 cmH20). The esophageal balloon was filled with 1 - 1.5 ml or air and positioned as described by Baydur et al. [16].
index (RVSWI) was calculated as the pressure gradient between the mean pulmonary arterial pressure (mPpa) and Pra (calculated as the mean over the systolic time period) multiplied by stroke volume (SI): RVSWI = ( m P p a - P r a ) . SI Right ventricular end diastolic volume index (RVEDVI) and right ventricular end systolic volume index ((RVESVI) were calculated by the following formulae: RVEDVI = SI/RVEF RVESVI = R V E D V I - SI
Measurements Data acquistion started after hemodynamic stabilization. The arterial and pulmonary arterial catheters were connected to pressure quartz transducers (Hewlett-Packard P 1290A Cupertino CA). Arterial pressure was continuously recorded, while pulmonary pressures and Pra were alternatively recorded through a stopcock system. The midaxillary line was taken as zero reference level for hemodynamic pressure measurements and pressures were read as electronic mean values at end-expiration. The ventilatory volume was determined electronically by integration of the flow signal. The Peso and Paw were measured continuously. All signals were recorded on an eight-channel strip chart recorder (Hewlett-Packard 7719 A) and on a personal computer via a 12-bit analog-digital converter at a sample rate of 50 Hz. Cardiac output was measured by the thermodilution technique (Edwards-Baxter REF-i Ejection Fraction/Cardiac Output Computer) using injections of 5 ml cold (
