Acta anaesth. scand. 1975, 19, 2 10-2 18
Evaluation of Impedance Cardiography as a Non-Invasive Means of Measuring Systolic Time Intervals and Cardiac Output J. P. RASMUSSEN, B. S0RENSEN and T. KANN Department of Anesthesia 11, Gentofte Hospital, Copenhagen, Denmark
Impedance cardiography was used for non-invasive determinations of systolic time intervals (STI) and cardiac output. The results were compared with simultaneously obtained invasive measurements of STI from central aortic pressure curves and of cardiac output using the dye-dilution technique. The study was performed on eight dogs during increasing halothane concentration. A close correlation was found between non-invasively and invasively measured left ventricular ejection time = LVET (7 = 0.986) and pre-cjection period = PEP (7 = 0.948). Measurements of cardiac output derived from changcs in thoracic impedance were determined 1 ) using a fixed value of p ( p = the resistivity of blood) and 2 ) using an individual value of p based on the actual hematocrits. When compared to cardiac outputs obtained by dye-dilution the correlation coefficients were r = 0.806 and r = 0.816, respectively. Impedance cardiography is a useful method of evaluating changes in cardiac output. The method permits simultaneous observations of changes in STI and cardiac output as an index of cardiac function. Received 17 January, accejted f o r publication 10 February 1975
The electrical impedance changes of the thorax associated with physiological activity have been used for non-invasive monitoring of respiration (BAKER& GEDDES1970), detection of thoracic fluid volume (VAK DE WATERet al. 1971), and measurements of cardiac output (KUBICEKet al. 1966). The stroke volume of the heart is related to the maximum rate of decrease in the electrical impedance occurring in the initial part of the cardiac cycle. Comparisons with stroke volume derived from the dye-dilution method have indicated a reasonable correlation between the two methods (KURICEK et al. 1966, BAKERet al. 1971, HILL& LOWE1973). Evaluation of myocardial function in man by the alterations in the temporal course of the cardiac cycle was introduced by KATZ& FEIL (1923) who established the method of simultaneous recording of heart sounds,
central aortic pulse tracing and electrocardiogram to define the different intervals in the cardiac systole. The total electrothe pre-ejection mechanical systole period (PEP) and the left ventricular ejection time (LVET) have been evaluated clinically (WEISSLER et al. 1968, WEISSLER & GARRARD 1971a, 1971b, FABIAN 1972a, 1972bj. Changes in the ratio between the pre-ejection period and the left ventricular ejection time (PEP/ LVET-ratio) have been shown to be a semiquantitative index of changes in stroke volume (~VEISSLERet al. 1969), while thc reciprocal value of the square of the preejection period ( 1/PEP2)is linearly correlated to peak ascending aortic blood flow acceleration (REITANet al. 1972). The purpose of this study was to evaluatc impedance cardiography - a non-invasivc method - as a means of measuring systolic
(as2),
NON-ISVASIVE MONITORING time intervals (STI) and cardiac output in the dog during halothane anesthesia. Halothane was chosen because of its well-known effect on the cardin-vascular system (EGER et al. 1970, EGERet al. 1971).
13Y IMPEDANCE CAKDIOGRAI’HY
211
inspired gas. Pao, was kept at 130+10 mmHg by varying FI,, in the NzO/Oz-mixture (8 Ijmin). Pa,,, Pacoz and pH were monitored continuously via an artificial external femoral arterial-vcnous shunt utilizing Radiometera equipment, Copenhagen (HENNINGSEN 1968). The dogs were kept heparinized throughout the study.
MATERIALS AKD METHODS Anesthesia and monitoring Procedures
Iiiifednnce cardiograflcy
Eight unpremedicated mongrel dogs weighing 12-20 kg (mean 16 kg) were anebtlietized with nitrous oxide 75% and oxygen after halothane induction via a Fluoteca Mark I1 vaporizer. Succinylcholine dichloride was injected intravenously, and after intubation, the dogs were ventilated with an intermittent positive pressure non-rebreathing system utilizing an Air-Shield@ Ventilator. Tidal volume (20 w/kg b.w.), respiratory rate (10/min) and inspiratory pressure (18-20 cmHZO) were kept constant. Pa,,, was maintained at 37-40 mmHg throughout the study by adding COZ to the
‘Transthoracic electrical impedance was measured by a four electrode impedancc cardiograph IFM/Minnesota Model 304A. Two conductive strip electrodes were placed around the neck and two around the abdomen at the level of the xiphoid process. A sinusoidal constant current (4 mA) at 100 kHz was applied to the outer two elcctrodes (Fig. 1). Potential changes reflecting impedance changes (AZ) accompanying cardiac activity were measured between the two inner electrodes. The first derivative of the AZ changes (dz/dt) was determined and recorded (Fig. 2).
Fig. 1. A schematic model of the methods employed. The mode of action of the impedancc cardiograph is demonstrated in block diagrams. AOP = aortic pressure curve; CVP = central venous pressure; I = electrical current; E = voltage; Z = impedance; ID = phonomicrophone; 2, = average thoracic impedance; AZ = impedance change; dz/dt = the first derivative of impedance change.
212
J. P. RASMUSSEN, B. SORENSEN A N D 1. KANN
Thoracic impedance stroke volume (SV) was calculated using the formula given by KUBICEK et al. (1 966) : SV = p x
(&)
x LVET x (dz/dt),i,,
where SV = stroke volume (cc) p = the electrical resistivity of blood at 100 kHz (ohm x cms) L = the mean distance between the two inner electrodes (cm) Z o = the basal thoracic impedance measured between the two inner electrodes (ohm) LVET = the left ventricular ejection time (s) (dz/dt),i, = dz/dt is the time rate of change of impedance during the cardiac cycle, and (dz/dt),,i, is the value a t the negative peak (ohmis).
jected through the catheter for eytimation of cardiac output Blood was drawn from the left femoral artery for dye-dilution curves obtained on a dyedensitometer and recorder developed by our biomedical engineering department. Total peripheral resistance (TPR) was calculated by dividing the difference bctween mean arterial arid central venous pressure by cardiac output.
(a).
Experimetital protocol
The experimental protocol consisted of a stabilization period of 60 minutes after the surgical proccdure. All dogs were then given increasing inspired halothane concentrations: 0.5-1 .O-1.5-2.0 vol.%. Thereafter, the investigation was continued on tlircc dogs (nos. 5, 7 and 8) where halothanc had first bcen discontinued for observation of recovery. Halothane was again administered 30 minutes later with an inspired concentration of 4.0 v01.x to observe sudden changes following a large concentration. The impedance SV was calculated first assuming a Each period with different inspired halothane convalue of p = 150 ohm x cms (an average value for the centrations lasted 30 min. Recordings of the STI and electrical resistivity of blood), and secondly using a n were performed every 10 minutes. All recordings were instantaneous specific value of p derived from a hematoobtained simultaneously on the Ultralette recorder crit resistivity C L I I ' V ~ (BAKERet al. 1971) based on the (frequency response of 1700 cycles per second) at a actual hematocrit read from a n Adams Readicrita paperspeed of 100 mm/s during endexpiratory apnea. centrifuge 059 1. Ten consecutive heart beats were analyzed and avcraged The aortic pressure curve (AOP) was recorded via a catheter with one endhole (Lehman no. 7, USCI@) in each recording. placed just outside the aortic valves. A Statham P-23 Db Strain gauge transducer was used. The transducer Statistics plus the aortic catheter have a 100 Hz flat amplitude Linear regression analysis and Wilcoxori-tests were used frequency response. for statistical analysis of the data. Lead I1 of the electrocardiogram (ECG), and the phonocardiogram (PCG) obtained from a n Elema" phonomicrophone ( E M T 2 1 B, amplifier E M T 28 B) placed at ictus cordis were recorded.
(i
RESULTS Systolic time interuals The following systolic time intervals were measured: 1 ) total electro-mechanical systole (QSZ) from the onset of the Q-wave of the ECG to the initial high frequency vibration of the second heart sound (S2), 2 ) left ventricular ejection time (LVET) 2a) from the upstroke of the AOP-curve to the dicrotic notch which correlates to Szof the PCG and 2b) from the zerocrossing of the dz/dt-curve before the negative peak (negative upward) to Sz of the PCG (Fig. 2). Preejection period (PEP) was calculated by subtracting LVET from QS,. PEP, PEP/LVET-ratio and l/PEPZ were all calculated from both the AOP-curve and the dz/dt-curve. Central D C I Z ~ U pressure, S cardiac outbut a d total peiipheral resistance Central venous pressure (CVP) was measurcd from the superior vena cava, and Cardio-green@ was in-
IVith increasing halothane concentrations, increases in PEP and PEP/LVET-ratio were observed, accompanied by a decrease in Q and arterial blood pressure (Table 1). Significant changes were observed in PEP and PEP/LVET-ratio (1'
Evaluation of Impedance Cardiography as a Non-Invasive Means of Measuring Systolic Time Intervals and Cardiac Output J. P. RASMUSSEN, B. S0RENSEN and T. KANN Department of Anesthesia 11, Gentofte Hospital, Copenhagen, Denmark
Impedance cardiography was used for non-invasive determinations of systolic time intervals (STI) and cardiac output. The results were compared with simultaneously obtained invasive measurements of STI from central aortic pressure curves and of cardiac output using the dye-dilution technique. The study was performed on eight dogs during increasing halothane concentration. A close correlation was found between non-invasively and invasively measured left ventricular ejection time = LVET (7 = 0.986) and pre-cjection period = PEP (7 = 0.948). Measurements of cardiac output derived from changcs in thoracic impedance were determined 1 ) using a fixed value of p ( p = the resistivity of blood) and 2 ) using an individual value of p based on the actual hematocrits. When compared to cardiac outputs obtained by dye-dilution the correlation coefficients were r = 0.806 and r = 0.816, respectively. Impedance cardiography is a useful method of evaluating changes in cardiac output. The method permits simultaneous observations of changes in STI and cardiac output as an index of cardiac function. Received 17 January, accejted f o r publication 10 February 1975
The electrical impedance changes of the thorax associated with physiological activity have been used for non-invasive monitoring of respiration (BAKER& GEDDES1970), detection of thoracic fluid volume (VAK DE WATERet al. 1971), and measurements of cardiac output (KUBICEKet al. 1966). The stroke volume of the heart is related to the maximum rate of decrease in the electrical impedance occurring in the initial part of the cardiac cycle. Comparisons with stroke volume derived from the dye-dilution method have indicated a reasonable correlation between the two methods (KURICEK et al. 1966, BAKERet al. 1971, HILL& LOWE1973). Evaluation of myocardial function in man by the alterations in the temporal course of the cardiac cycle was introduced by KATZ& FEIL (1923) who established the method of simultaneous recording of heart sounds,
central aortic pulse tracing and electrocardiogram to define the different intervals in the cardiac systole. The total electrothe pre-ejection mechanical systole period (PEP) and the left ventricular ejection time (LVET) have been evaluated clinically (WEISSLER et al. 1968, WEISSLER & GARRARD 1971a, 1971b, FABIAN 1972a, 1972bj. Changes in the ratio between the pre-ejection period and the left ventricular ejection time (PEP/ LVET-ratio) have been shown to be a semiquantitative index of changes in stroke volume (~VEISSLERet al. 1969), while thc reciprocal value of the square of the preejection period ( 1/PEP2)is linearly correlated to peak ascending aortic blood flow acceleration (REITANet al. 1972). The purpose of this study was to evaluatc impedance cardiography - a non-invasivc method - as a means of measuring systolic
(as2),
NON-ISVASIVE MONITORING time intervals (STI) and cardiac output in the dog during halothane anesthesia. Halothane was chosen because of its well-known effect on the cardin-vascular system (EGER et al. 1970, EGERet al. 1971).
13Y IMPEDANCE CAKDIOGRAI’HY
211
inspired gas. Pao, was kept at 130+10 mmHg by varying FI,, in the NzO/Oz-mixture (8 Ijmin). Pa,,, Pacoz and pH were monitored continuously via an artificial external femoral arterial-vcnous shunt utilizing Radiometera equipment, Copenhagen (HENNINGSEN 1968). The dogs were kept heparinized throughout the study.
MATERIALS AKD METHODS Anesthesia and monitoring Procedures
Iiiifednnce cardiograflcy
Eight unpremedicated mongrel dogs weighing 12-20 kg (mean 16 kg) were anebtlietized with nitrous oxide 75% and oxygen after halothane induction via a Fluoteca Mark I1 vaporizer. Succinylcholine dichloride was injected intravenously, and after intubation, the dogs were ventilated with an intermittent positive pressure non-rebreathing system utilizing an Air-Shield@ Ventilator. Tidal volume (20 w/kg b.w.), respiratory rate (10/min) and inspiratory pressure (18-20 cmHZO) were kept constant. Pa,,, was maintained at 37-40 mmHg throughout the study by adding COZ to the
‘Transthoracic electrical impedance was measured by a four electrode impedancc cardiograph IFM/Minnesota Model 304A. Two conductive strip electrodes were placed around the neck and two around the abdomen at the level of the xiphoid process. A sinusoidal constant current (4 mA) at 100 kHz was applied to the outer two elcctrodes (Fig. 1). Potential changes reflecting impedance changes (AZ) accompanying cardiac activity were measured between the two inner electrodes. The first derivative of the AZ changes (dz/dt) was determined and recorded (Fig. 2).
Fig. 1. A schematic model of the methods employed. The mode of action of the impedancc cardiograph is demonstrated in block diagrams. AOP = aortic pressure curve; CVP = central venous pressure; I = electrical current; E = voltage; Z = impedance; ID = phonomicrophone; 2, = average thoracic impedance; AZ = impedance change; dz/dt = the first derivative of impedance change.
212
J. P. RASMUSSEN, B. SORENSEN A N D 1. KANN
Thoracic impedance stroke volume (SV) was calculated using the formula given by KUBICEK et al. (1 966) : SV = p x
(&)
x LVET x (dz/dt),i,,
where SV = stroke volume (cc) p = the electrical resistivity of blood at 100 kHz (ohm x cms) L = the mean distance between the two inner electrodes (cm) Z o = the basal thoracic impedance measured between the two inner electrodes (ohm) LVET = the left ventricular ejection time (s) (dz/dt),i, = dz/dt is the time rate of change of impedance during the cardiac cycle, and (dz/dt),,i, is the value a t the negative peak (ohmis).
jected through the catheter for eytimation of cardiac output Blood was drawn from the left femoral artery for dye-dilution curves obtained on a dyedensitometer and recorder developed by our biomedical engineering department. Total peripheral resistance (TPR) was calculated by dividing the difference bctween mean arterial arid central venous pressure by cardiac output.
(a).
Experimetital protocol
The experimental protocol consisted of a stabilization period of 60 minutes after the surgical proccdure. All dogs were then given increasing inspired halothane concentrations: 0.5-1 .O-1.5-2.0 vol.%. Thereafter, the investigation was continued on tlircc dogs (nos. 5, 7 and 8) where halothanc had first bcen discontinued for observation of recovery. Halothane was again administered 30 minutes later with an inspired concentration of 4.0 v01.x to observe sudden changes following a large concentration. The impedance SV was calculated first assuming a Each period with different inspired halothane convalue of p = 150 ohm x cms (an average value for the centrations lasted 30 min. Recordings of the STI and electrical resistivity of blood), and secondly using a n were performed every 10 minutes. All recordings were instantaneous specific value of p derived from a hematoobtained simultaneously on the Ultralette recorder crit resistivity C L I I ' V ~ (BAKERet al. 1971) based on the (frequency response of 1700 cycles per second) at a actual hematocrit read from a n Adams Readicrita paperspeed of 100 mm/s during endexpiratory apnea. centrifuge 059 1. Ten consecutive heart beats were analyzed and avcraged The aortic pressure curve (AOP) was recorded via a catheter with one endhole (Lehman no. 7, USCI@) in each recording. placed just outside the aortic valves. A Statham P-23 Db Strain gauge transducer was used. The transducer Statistics plus the aortic catheter have a 100 Hz flat amplitude Linear regression analysis and Wilcoxori-tests were used frequency response. for statistical analysis of the data. Lead I1 of the electrocardiogram (ECG), and the phonocardiogram (PCG) obtained from a n Elema" phonomicrophone ( E M T 2 1 B, amplifier E M T 28 B) placed at ictus cordis were recorded.
(i
RESULTS Systolic time interuals The following systolic time intervals were measured: 1 ) total electro-mechanical systole (QSZ) from the onset of the Q-wave of the ECG to the initial high frequency vibration of the second heart sound (S2), 2 ) left ventricular ejection time (LVET) 2a) from the upstroke of the AOP-curve to the dicrotic notch which correlates to Szof the PCG and 2b) from the zerocrossing of the dz/dt-curve before the negative peak (negative upward) to Sz of the PCG (Fig. 2). Preejection period (PEP) was calculated by subtracting LVET from QS,. PEP, PEP/LVET-ratio and l/PEPZ were all calculated from both the AOP-curve and the dz/dt-curve. Central D C I Z ~ U pressure, S cardiac outbut a d total peiipheral resistance Central venous pressure (CVP) was measurcd from the superior vena cava, and Cardio-green@ was in-
IVith increasing halothane concentrations, increases in PEP and PEP/LVET-ratio were observed, accompanied by a decrease in Q and arterial blood pressure (Table 1). Significant changes were observed in PEP and PEP/LVET-ratio (1'
