• View in gallery

    Bland-Altman analysis of sCO values measured concurrently by use of the LIDCO and PICCO methods in 6 anesthetized neonatal foals. Values were measured during a low CO state (black circles), an intermediate CO state (white circles), and a high CO state (white squares) in each foal; however, the PICCO method failed to provide values in 4 of 18 attempts.

  • View in gallery

    Bland-Altman analysis of sCO values measured concurrently by use of the LIDCO and PULSECO methods in 6 anesthetized neonatal foals during various CO states. The mean bias (solid horizontal line) and upper and lower limits of agreements (defined as the mean plus [1.96 × SD] and the mean minus [1.96 × SD], respectively; dashed lines) are indicated. See Figure 1 for remainder of key.

  • 1.

    Valverde A, Giguère S, Sanchez LC, et al. Effects of dobutamine, norepinephrine, and vasopressin on cardiovascular function in anesthetized neonatal foals with induced hypotension. Am J Vet Res 2006;67:17301737.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 2.

    Valverde A, Giguère S, Morey TE, et al. Comparison of noninvasive cardiac output measured by use of partial carbon dioxide rebreathing or the lithium dilution method in anesthetized foals. Am J Vet Res 2007;68:141147.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 3.

    Tibby SM, Murdoch IA. Monitoring cardiac function in intensive care. Arch Dis Child 2003;88:4652.

  • 4.

    Corley KT, Donaldson LL, Furr MO. Comparison of lithium dilution and thermodilution cardiac output measurements in anaesthetised neonatal foals. Equine Vet J 2002;34:598601.

    • Search Google Scholar
    • Export Citation
  • 5.

    Connors AF Jr, Speroff T, Dawson NV, et al. The effectiveness of right heart catheterization in the initial care of critically ill patients. SUPPORT Investigators. JAMA 1996;276:889897.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6.

    Connors AF Jr, Castele RJ, Farhat NZ, et al. Complications of right heart catheterization. A prospective autopsy study. Chest 1985;88:567572.

  • 7.

    Sandham JD, Hull RD, Brant RF, et al. A randomized, controlled trial of the use of pulmonary-artery catheters in high-risk surgical patients. N Engl J Med 2003;348:514.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8.

    Linton RA, Young LE, Marlin DJ, et al. Cardiac output measured by lithium dilution, thermodilution, and transesophageal Doppler echocardiography in anesthetized horses. Am J Vet Res 2000;61:731737.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9.

    Corley KT, Donaldson LL, Durando MM, et al. Cardiac output technologies with special reference to the horse. J Vet Intern Med 2003;17:262272.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10.

    Giguère S, Bucki E, Adin DB, et al. Cardiac output measurement by partial carbon dioxide rebreathing, 2-dimensional echocardiography, and lithium-dilution method in anesthetized neonatal foals. J Vet Intern Med 2005;19:737743.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11.

    Mason DJ, O'Grady M, Woods JP, et al. Effect of background serum lithium concentrations on the accuracy of lithium dilution cardiac output determination in dogs. Am J Vet Res 2002;63:10481052.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12.

    Beaulieu JM, Sotnikova TD, Yao WD, et al. Lithium antagonizes dopamine-dependent behaviors mediated by an AKT/glycogen synthase kinase 3 signaling cascade. Proc Natl Acad Sci U S A 2004;101:50995104.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13.

    Cooper ES, Muir WW. Continuous cardiac output monitoring via arterial pressure waveform analysis following severe hemorrhagic shock in dogs. Crit Care Med 2007;35:17241729.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14.

    Chen HC, Sinclair MD, Dyson DH, et al. Comparison of arterial pressure waveform analysis with the lithium dilution technique to monitor cardiac output in anesthetized dogs. Am J Vet Res 2005;66:14301436.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15.

    van Lieshout JJ, Wesseling KH. Continuous cardiac output by pulse contour analysis? Br J Anaesth 2001;86:467469.

  • 16.

    Linton NW, Linton RA. Estimation of changes in cardiac output from the arterial blood pressure waveform in the upper limb. Br J Anaesth 2001;86:486496.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17.

    Orme RM, Pigott DW, Mihm FG. Measurement of cardiac output by transpulmonary arterial thermodilution using a long radial artery catheter. A comparison with intermittent pulmonary artery thermodilution. Anaesthesia 2004;59:590594.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18.

    de Wilde RB, Schreuder JJ, van den Berg PC, et al. An evaluation of cardiac output by five arterial pulse contour techniques during cardiac surgery. Anaesthesia 2007;62:760768.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 19.

    Fakler U, Pauli C, Balling G, et al. Cardiac index monitoring by pulse contour analysis and thermodilution after pediatric cardiac surgery. J Thorac Cardiovasc Surg 2007;133:224228.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20.

    Halvorsen PS, Espinoza A, Lundblad R, et al. Agreement between PiCCO pulse-contour analysis, pulmonal artery thermodilution and transthoracic thermodilution during off-pump coronary artery by-pass surgery. Acta Anaesthesiol Scand 2006;50:10501057.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21.

    Jonas MM, Tanser SJ. Lithium dilution measurement of cardiac output and arterial pulse waveform analysis: an indicator dilution calibrated beat-by-beat system for continuous estimation of cardiac output. Curr Opin Crit Care 2002;8:257261.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 22.

    Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res 1999;8:135160.

  • 23.

    Shoemaker WC, Wo CC, Bishop MH, et al. Multicenter trial of a new thoracic electrical bioimpedance device for cardiac output estimation. Crit Care Med 1994;22:19071912.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 24.

    Lin LI-K. A concordance correlation coefficient to evaluate reproducibility. Biometrics 1989;45:255268.

  • 25.

    Lin LI-K. A note on the concordance correlation coefficient. Biometrics 2000;56:324325.

  • 26.

    de Wilde RB, Breukers RB, van den Berg PC, et al. Monitoring cardiac output using the femoral and radial arterial pressure waveform. Anaesthesia 2006;61:743746.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 27.

    Orme RM, Pigott DW. Monitoring cardiac output from the radial artery pressure waveform. Anaesthesia 2007;62:9091.

  • 28.

    Critchley LA, Critchley JA. A meta-analysis of studies using bias and precision statistics to compare cardiac output measurement techniques. J Clin Monit Comput 1999;15:8591.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 29.

    Hallowell GD, Corley KT. Use of lithium dilution and pulse contour analysis cardiac output determination in anaesthetized horses: a clinical evaluation. Vet Anaesth Analg 2005;32:201211.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30.

    Costa MG, Della Rocca G, Chiarandini P, et al. Continuous and intermittent cardiac output measurement in hyperdynamic conditions: pulmonary artery catheter vs. lithium dilution technique. Intensive Care Med 2008;34:257263.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 31.

    Missant C, Rex S, Wouters PF. Accuracy of cardiac output measurements with pulse contour analysis (PulseCO) and Doppler echocardiography during off-pump coronary artery bypass grafting. Eur J Anaesthesiol 2008;25:243248.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32.

    Yamashita K, Nishiyama T, Yokoyama T, et al. Effects of vasodilation on cardiac output measured by PulseCO. J Clin Monit Comput 2007;21:335339.

Advertisement

Determination of cardiac output in anesthetized neonatal foals by use of two pulse wave analysis methods

View More View Less
  • 1 Department of Large Animal Clinical Sciences, College of Veterinary Medicine, University of Florida, Gainesville, FL 32610.
  • | 2 Department of Large Animal Clinical Sciences, College of Veterinary Medicine, University of Florida, Gainesville, FL 32610.
  • | 3 Department of Large Animal Clinical Sciences, College of Veterinary Medicine, University of Florida, Gainesville, FL 32610.
  • | 4 Department of Clinical Studies, Ontario Veterinary College, University of Guelph, Guelph, ON N1G 2W1, Canada.
  • | 5 Department of Large Animal Clinical Sciences, College of Veterinary Medicine, University of Florida, Gainesville, FL 32610.
  • | 6 Department of Large Animal Clinical Sciences, College of Veterinary Medicine, University of Florida, Gainesville, FL 32610.

Abstract

Objective—To compare cardiac output (CO) measured by lithium arterial pressure waveform analysis (PULSECO) and CO measured by transpulmonary pulse contour analysis (PICCO) in anesthetized foals, with CO measured by use of lithium dilution (LIDCO) considered the criterion-referenced standard.

Sample Population—6 neonatal (1- to 4-day-old) foals that weighed 38 to 45 kg.

Procedures—Foals were anesthetized and instrumented to measure direct blood pressure, heart rate, arterial blood gases, and CO. The CO was measured by use of PULSECO, PICCO, and LIDCO techniques. Measurements were converted to specific CO (sCO) values for statistical analysis. Measurements were obtained during low, intermediate, and high CO states.

Results—sCO ranged from 75.5 to 310 mL/kg/min. Mean ± SD PICCO bias varied significantly among CO states and was −51.9 ± 23.1 mL/kg/min, 20.0 ± 19.5 mL/kg/min, and 87.2 ± 19.5 mL/kg/min at low, intermediate, and high CO states, respectively. Mean PULSECO bias (11.0 ± 37.5 mL/kg/min) was significantly lower than that of PICCO and did not vary among CO states. Concordance correlation coefficient between LIDCO and PULSECO was significantly greater than that between LIDCO and PICCO. The proportion of observations with a relative bias < ± 30% was significantly lower with the PULSECO method than with the PICCO method.

Conclusions and Clinical Relevance—Values for the PULSECO method were more reproducible and agreed better with values for the LIDCO method than did values for the PICCO method and were able to more accurately monitor changes in CO in anesthetized newborn foals.

Abstract

Objective—To compare cardiac output (CO) measured by lithium arterial pressure waveform analysis (PULSECO) and CO measured by transpulmonary pulse contour analysis (PICCO) in anesthetized foals, with CO measured by use of lithium dilution (LIDCO) considered the criterion-referenced standard.

Sample Population—6 neonatal (1- to 4-day-old) foals that weighed 38 to 45 kg.

Procedures—Foals were anesthetized and instrumented to measure direct blood pressure, heart rate, arterial blood gases, and CO. The CO was measured by use of PULSECO, PICCO, and LIDCO techniques. Measurements were converted to specific CO (sCO) values for statistical analysis. Measurements were obtained during low, intermediate, and high CO states.

Results—sCO ranged from 75.5 to 310 mL/kg/min. Mean ± SD PICCO bias varied significantly among CO states and was −51.9 ± 23.1 mL/kg/min, 20.0 ± 19.5 mL/kg/min, and 87.2 ± 19.5 mL/kg/min at low, intermediate, and high CO states, respectively. Mean PULSECO bias (11.0 ± 37.5 mL/kg/min) was significantly lower than that of PICCO and did not vary among CO states. Concordance correlation coefficient between LIDCO and PULSECO was significantly greater than that between LIDCO and PICCO. The proportion of observations with a relative bias < ± 30% was significantly lower with the PULSECO method than with the PICCO method.

Conclusions and Clinical Relevance—Values for the PULSECO method were more reproducible and agreed better with values for the LIDCO method than did values for the PICCO method and were able to more accurately monitor changes in CO in anesthetized newborn foals.

Cardiac output is defined as the volume of blood ejected by the heart per unit of time. Cardiac output, blood hemoglobin concentration, and oxygen saturation of hemoglobin are major factors in determining global tissue oxygen delivery and consumption.1 Although CO is the best available method for assessing cardiovascular function, it is not routinely monitored because of the invasive nature and technical challenges of many currently used methods.2

The pulmonary thermodilution method has been accepted as the criterion-referenced standard for CO monitoring for many years.3,4 This technique requires catheterization of the pulmonary artery, which is not commonly performed in foals.5 Pulmonary artery catheterization increases mortality and morbidity rates in humans,6,7 and it can result in endocardial lesions in the right side of the heart and pulmonary artery in horses.8,a

Values of CO measured via the LIDCO method compare favorably with results obtained with thermodilution, and LIDCO is currently the method most commonly used in anesthetized foals.4,9,10 Drawbacks associated with repeated determination of CO via the LIDCO method include blood loss and accumulation of the indicator (lithium) in the body with the potential for undesirable effects.11,12 Furthermore, LIDCO only provides CO at a single time point, which is not ideal for monitoring critically ill patients with a labile cardiovascular status.13

Thus, there has been growing interest in the use of less invasive techniques for continuous measurement of CO, such as application of the Fick principle, esophageal Doppler ultrasonography, thoracic bioimpedance, and arterial pulse wave analysis.10 Deriving CO on the basis of the arterial pressure uses algorithms that correlate the area under the arterial pressure curve with stroke volume and CO.14,15 However, multiple factors, including changes in aortic compliance, reflected pressure waves, and waveform damping, prevent this from being a straightforward relationship.13,14

Two new hemodynamic monitors (ie, PULSECO and PICCO) offer minimally invasive, continuous, beat-to-beat CO determination.13,14 Either technique would potentially provide a good method for continuous CO monitoring in anesthetized or awake foals.

The PULSECO method uses pulse power analysis to determine CO from characteristics of the entire arterial pressure waveform and is calibrated by LIDCO for continuous beat-to-beat CO.16 The signal used for PULSECO is lithium chloride. In contrast, the PICCO method uses pulse contour analysis in accordance with a modified version of the algorithm of Wesseling17 that calculates stroke volume on a beat-to-beat basis by dividing the area under the systolic time curve (area before the dicrotic notch) by the aortic impedance; the aortic impedance is derived from transpulmonary aortic thermodilution.17 The signal used for PICCO is cold saline (0.9% NaCl) solution. Both techniques are limited by arrhythmias. In addition, PULSECO is limited by severe vasoconstriction and aortic valve regurgitation, whereas PICCO is limited by valvular insufficiency and rapid changes in body temperature. Although both techniques provide continuous CO monitoring, PICCO is also able to provide valuable data regarding a patient's preload, afterload, and extravascular lung water.

Investigators in some experimental and clinical studies14,18–20 have compared PULSECO or PICCO to lithium and thermodilution indicator methods in humans and other animals. To the authors' knowledge, no one has conducted a study to concurrently compare both pulse analysis CO methods in foals. The objective of the study reported here was to analyze the agreement between sCO determined by LIDCO and either PULSECO or PICCO in anesthetized neonatal foals.

Materials and Methods

Animals—Six neonatal foals between 1 and 3 days of age and weighing between 38 and 45 kg were used in the study. Each foal had a normal birth and was determined to be healthy on the basis of results of physical examination. The study was approved by the University of Florida Institutional Animal Care and Use Committee (protocol No. E-619).

Anesthesia and instrumentation—Each foal was sedated by administration of diazepamb (0.1 mg/kg, IV), anesthetized by administration of ketamine hydrochloridec (2.2 mg/kg, IV), and positioned in left lateral recumbency. A nasotracheal tube (internal diameter, 9 to 12 mm) was placed for administration of isofluraned vaporized in oxygen via a circle system with an oxygen flow of 2 to 3 L/kg. Foals were mechanically ventilated at a rate of 8 to 10 breaths/min and a tidal volume of 10 to 15 mL/kg. The PETCO2 and end-tidal isoflurane concentrations were monitored with an infrared gas analyzer.e A 7-F, 30-cm, double-lumen catheterf was inserted in a jugular vein of each foal; the tip of the catheter was directed into the right atrium. Values for systolic arterial pressure, diastolic arterial pressure, and mean arterial pressure were obtained by use of a 20-gauge, 1.88-inch catheterg placed in the right metatarsal artery and attached to an electronic pressure transducer positioned and calibrated to a zero value at the level of the sternal manubrium. For PICCO measurements, a 4-F, 50-cm, PICCO arterial catheterh was also placed in a metatarsal artery and advanced toward the heart. Each foal was instrumented with ECG leads for continuous monitoring of heart rate and rhythm. Normovolemia and normoglycemia were maintained by IV administration of a balanced electrolyte solution with 2.5% dextrosei at a rate of 4 mL/kg/h.

Measurement of CO via LIDCO—Blood hemoglobin and serum sodium concentrations required for CO measurement were determined by use of a blood gas analyzerj immediately before performing the initial measurement. Lithium chloridek (0.003 mmol/kg) was injected via the venous catheter, and arterial blood was withdrawn through a lithium sensorl placed between the side port of the 3-way stopcock and the metatarsal arterial catheter. Blood was collected through a peristaltic pumpm at a flow rate of 4 mL/min across the sensor. Cardiac output was automatically derived from the concentration versus time curve by use of the system software.n

Measurement of CO via PULSECO—The PULSECO monitoro was connected to a direct blood pressure monitor. The analogue arterial pressure tracing underwent a 3-step transformation by the PULSECO algorithm to derive an estimation of CO.21 First, a nominal stroke volume was determined; CO was then calculated by multiplying this stroke volume by the heart rate. A calibration factor was determined by use of the LIDCO method.14 After the 1-point calibration, the PULSECO monitor displayed a beat-by-beat value for CO. Only 1 calibration with the LIDCO was performed at the beginning of the experiment (ie, baseline value). All the following LIDCO measurements were not entered in the PULSECO computer. The PULSECO software was modified to allow us to use lithium dilution and not recalibrate the PULSECO to those values. This allowed evaluation of how the PULSECO monitor would handle various CO states without recalibration.

Measurement of CO via PICCO—A 50-cm PICCO catheter was placed in the proximal metatarsal artery and advanced by use of a guide wire and dilator so that the tip was located within the cranial tibial or caudal femoral arteries. Cardiac output was calculated on the basis of the area under the systolic time curve (area before the dicrotic notch) and displayed continuously. Calibration baseline CO was determined via aortic transpulmonary thermodilution. Briefly, a 20-mL bolus of cold (< 8°C) saline solution was injected into the distal port of the central venous catheter. Alteration in blood temperature was detected by a thermistor located on the distal end of the arterial catheter. Cardiac output was automatically calculated with system software by use of a modified Stewart-Hamilton equation.17 Transpulmonary thermodilution calibration was performed only at the beginning of the experiments (ie, baseline).

Experimental design—After each foal was instrumented, baseline LIDCO and transpulmonary thermodilution CO measurements were obtained for calibration of the PULSECO method and PICCO method, respectively. Three additional CO states were evaluated. For each CO state, foals were maintained at that hemodynamic plane for at least 10 minutes prior to data collection. The low CO state was obtained by maintaining anesthesia at an end-tidal isoflurane concentration of 1.3% to 2.1%. For the intermediate CO state, end-tidal isoflurane concentration was decreased to 0.9% to 1.4% and dobutaminep (1 to 3 μg/kg/min) was administered by use of an infusion pump.q The high CO state was attained by further changing the depth of anesthesia to achieve an end-tidal isoflurane concentration of 0.8% to 1.0% and increasing the rate of dobutamine infusion (3 to 6 μg/kg/min). For each CO state, measurements were performed in duplicate; there was an interval of at least 3 minutes between duplicate measurements, and the mean value for duplicate measurements was calculated and used for comparisons. When any 2 CO measurements for a given method varied by > 20%, a third measurement was obtained, and the measurement with variation > 20% from the resulting mean for the 3 measurements was eliminated. Data were expressed as sCO, which was obtained by dividing CO by body weight.

Statistical analysis—Baseline LIDCO values were compared with concurrent baseline transpulmonary thermodilution values by use of a paired t test. Agreement between PICCO or PULSECO and the reference method (ie, LIDCO) was determined by use of the method of Bland and Altman.22 For each observation, bias was calculated as follows: (sCOLIDCO – sCOPICCO [or sCOPULSECO]), where sCOLIDCO and sCOPICCO or sCOPULSECO were the sCO values at a given time measured concurrently by use of LIDCO, PICCO, and PULSECO, respectively. For each observation, bias was also expressed as a percentage of the mean sCO (relative bias). Relative bias was calculated as follows: (sCOLIDCO – sCOPICCO [or sCOPULSECO])/(0.5 × [sCOLIDCO + sCOPICCO {or sCOPULSECO}]) × 100.23 For PICCO and PULSECO, a positive bias or relative bias reflected underestimation of the LIDCO-derived CO, whereas a negative value indicated overestimation of the LiDCO-derived CO. The limits of agreement were reported as bias (or relative bias) ± (1.96 × SD).

Normality of the data and equality of variances were assessed by use of the Kolmogorov-Smirnov and Levene tests, respectively. A general linear model with repeated observations was conducted to assess the effect of the method of measurement and CO state (low, intermediate, and high) on bias.

A similar general linear model with repeated observations was conducted to assess differences in the absolute value of the bias between methods. For effects that were significant for an overall F test, pairwise comparisons were made by use of the Student-Newman-Keuls method. In a study23 in which investigators assessed CO methods in humans, it was determined by combining the errors of both the test and reference method that acceptance of a new method for measurement of CO should rely on limits of agreement of up to approximately ± 30%. As a result, the proportion of observations with a relative bias ≥ 30% was compared between the PICCO and PULSECO methods by use of the Fisher exact test. Data were also compared by the use of Lin concordance correlation, which compares 2 techniques measuring the same variable without the inherent bias of establishing a criterion-referenced standard.24,25 The concordance correlation coefficient was calculated as follows: Rc = R × Cb, where Rc is the concordance correlation coefficient; R is the Pearson correlation coefficient, which measures the amount that each observation deviates from the best-fit line (ie, precision); and Cb is a bias correction factor, which measures the amount that the best-fit line deviates from the line with a slope of 45° that passes through the origin (ie, accuracy).24,25 A Rc of 1 indicates perfect concordance. For all analyses, values of P < 0.05 were considered significant.

Results

Mean ± SD baseline lithium CO value (7.6 ± 0.9 L/min) used to calibrate the PULSECO method did not differ significantly (P = 0.254) from the mean baseline transpulmonary thermodilution CO value (9.3 ± 2.3 L/min) used to calibrate the PICCO method. Eighteen pairs of measurements were subsequently obtained from the 6 foals. Values for LIDCO ranged between 4.0 and 14.0 L/min (mean ± SD, 8.3 ± 3.4 L/min), which resulted in sCO values between 75.5 and 310.0 mL/kg/min (mean, 190.0 ± 75.0 mL/kg/min). The mean LIDCO sCO was 101.1 ± 17.7 mL/kg/min for the low CO state, 182.0 ± 15.0 mL/kg/min for the intermediate CO state, and 274.0 ± 26.3 mL/kg/min for the high CO state. Total amount of lithium chloride administered per foal was 0.028 ± 0.01 mmol/kg.

The PICCO method failed to provide data in 4 of 18 attempts. There was a significant effect of CO state on bias (P < 0.001) and relative bias (P = 0.003) for the PICCO method. Mean ± SD bias of the PICCO method was −51.9 ± 23.1 mL/kg/min, 20.0 ± 19.5 mL/kg/min, and 87.2 ± 19.5 mL/kg/min at the low, intermediate, and high CO states, respectively (Figure 1). The PULSECO method provided data at each attempt. There was not a significant (power, 89%) effect of CO state on bias and relative bias for the PULSECO method. Mean ± SD bias of the PULSECO method was −13.1 ± 20.3 mL/kg/min, 18.7 ± 16.6 mL/kg/min, and 22.8 ± 19.3 mL/kg/min for the low, intermediate, and high CO states, respectively (Figure 2). Mean bias for the PULSECO method was 11.0 ± 37.5 mL/kg/min (limits of agreement, −62.5 mL/kg/min). Mean relative bias for the PULSECO method was 3.9 ± 17.8% (limits of agreements, −34.9% to 42.8%).

Figure 1—
Figure 1—

Bland-Altman analysis of sCO values measured concurrently by use of the LIDCO and PICCO methods in 6 anesthetized neonatal foals. Values were measured during a low CO state (black circles), an intermediate CO state (white circles), and a high CO state (white squares) in each foal; however, the PICCO method failed to provide values in 4 of 18 attempts.

Citation: American Journal of Veterinary Research 70, 3; 10.2460/ajvr.70.3.334

Figure 2—
Figure 2—

Bland-Altman analysis of sCO values measured concurrently by use of the LIDCO and PULSECO methods in 6 anesthetized neonatal foals during various CO states. The mean bias (solid horizontal line) and upper and lower limits of agreements (defined as the mean plus [1.96 × SD] and the mean minus [1.96 × SD], respectively; dashed lines) are indicated. See Figure 1 for remainder of key.

Citation: American Journal of Veterinary Research 70, 3; 10.2460/ajvr.70.3.334

The relative performance of each method was initially assessed by comparing the absolute value of their bias. Use of an ANOVA revealed a significant (P = 0.012) effect of method on sCO measurement, but there was not a significant effect of CO state (P = 0.157) or a significant interaction between CO state and method (P = 0.927). The absolute value of the bias of the PULSECO method was significantly lower than that of the PICCO method (Table 1). The concordance correlation coefficient between the LIDCO and PULSECO methods was significantly (P = 0.003) greater than that between the LIDCO and PICCO methods. In addition, the proportion of observations with a relative bias <± 30% was significantly (P = 0.023) lower with the PULSECO method than with the PICCO method.

Table 1—

Comparison of the PULSECO and PICCO methods for measurement of CO in 6 anesthetized neonatal foals.

CO methodMean ± SD absolute value of bias (mL/kg/min)Concordance correlation coefficient*Proportion of observations with relative bias < ± 30%
PULSECO37.8 ± 21.8a0.849 (0.654 to 0.939)a2 of 18a
PICCO63.5 ± 34.3b0.080 (-0.392 to 0.520)b8of14b

Values were measured by use of each method during low, intermediate, and high CO states in each foal; however, the PICCO method failed to provide values in 4 of 18 attempts.

Values in parentheses represent 95% confidence intervals.

Within a column, values with different superscript letters differ significantly (P < 0.05).

Discussion

In the study reported here, we evaluated the bias, precision, and accuracy of 2 pulse wave analysis techniques by simultaneous comparison of their sCO values against an accepted standard method (ie, LIDCO) in foals. This study was able to compare values for a wide physiologic range of sCO values.

To our knowledge, this is the first report of the use of the PICCO method in neonatal foals. The manufacturer recommends that the PICCO arterial catheter be placed in the femoral artery with the catheter tip located in the aorta (central catheterization) to detect central aortic blood pressure. Arterial central catheterization is not commonly performed in veterinary clinical settings, and the fact that the PICCO method requires femoral catheterization makes it less clinically applicable. We intended to use both CO pulse analysis methods in the more readily accessible proximal portion of a metatarsal artery to simulate clinical scenarios. Investigators in a recent study26 determined the interchangeability of femoral artery pressure and pressure measurements obtained from a peripheral artery as the input for the PICCO method. In humans, CO measured with the PICCO method by use of a 50-cm catheter in a peripheral artery was interchangeable with pulmonary thermodilution CO.17,26 Therefore, we deemed it appropriate to use pressure in a peripheral artery as an input for the PICCO method. In some cases in our study, the PICCO method failed to detect the bolus of cold saline solution and calibration was not achieved. Calibration may have been more successful if the thermistor-tipped arterial catheter had been placed more centrally (in the proximal femoral artery or aorta).17,27

Relative bias of the PICCO method varied significantly with the CO state. It greatly overestimated measurements in low CO states and underestimated CO in high CO states. The PICCO method would be less useful clinically because the CO state in a given patient is unknown.10 Change in the shape of the arterial waveform may explain the variability in agreement between pulse wave analysis and the LIDCO method. During the low CO state, the PICCO method had more difficulty detecting the dicrotic notch and calculating the area under the systolic time curve.

Both PULSECO and PICCO were calibrated only once at the beginning of the experiment, but because of the inherent characteristics of the equipment, the 2 techniques were calibrated by use of different methods (LIDCO vs transpulmonary thermodilution). The PULSECO software was modified to allow us to measure multiple LIDCO values without automatic recalibration. Unfortunately, the PICCO software could not be modified and would have automatically performed a recalibration for every thermodilution. Therefore, subsequent thermodilution CO values could not be obtained throughout the experiment. Furthermore, the PICCO equipment only accepted internal calibration with its own transpulmonary thermodilution system. Consequently, it was impossible to use the same method to calibrate both monitors. Baseline CO measured by use of transpulmonary thermodilution was not significantly different from that obtained by use of the LIDCO method. Thus, it is unlikely that the choice of the method used as a criterion-referenced standard (LIDCO) would have influenced the results of the study reported here.

In our study, the PULSECO monitor provided good directional tracking and good agreement with that for the reference method. In addition, the PULSECO method had a significantly lower proportion of observations with a relative bias <± 30% when compared with the PICCO method. Consequently, the PULSECO method can be used interchangeably with the lithium dilution method.10,28 Results for the PULSECO method are in agreement with other CO measurements obtained in horses. The mean bias and relative bias between the PULSECO and LIDCO methods in the study reported here were similar to those recorded in other studies in which investigators compared different CO techniques in horses. For example, when comparing the LIDCO method to the bullet transthoracic echocardiographic method for measurement of CO in neonatal foals, a mean ± SD relative bias of 4.2 ± 20.9% has been obtained.10 In another study,29 investigators reported a relative bias of 9% when comparing values for the LIDCO and PULSECO methods in mature horses. In humans, the PULSECO method has good agreement, compared with results for the LIDCO and pulmonary thermodilution methods.30,31 In contrast, the PULSECO method is less accurate in dogs, especially when the hemodynamic conditions change from those of the initial calibration.14,32

The sample size of our study was rather small (6 foals). We controlled variables that could have affected the hemodynamic and CNS status of the foals to minimize their effects on our results. Hemoglobin concentrations, which are important for LIDCO determinations, varied minimally despite the blood loss invariably associated with this technique. Blood loss associated with withdrawal of arterial blood for the LIDCO method was minimal and estimated at < 2.5% of the total blood volume. Similarly, the total mean cumulative dose of lithium administered was low and should not have affected our results.4,11

Analysis of the results of the study reported here indicated that the PICCO and PULSECO methods were both able to monitor CO changes in the same direction as detected by use of the LIDCO method (ie, pulse wave–derived CO increases as CO increases and vice versa). However, the PULSECO method was significantly more reproducible, with less variability and better agreement with the criterion-referenced method, than was the PICCO method in anesthetized neonatal foals.

Abbreviations

CO

Cardiac output

LIDCO

Cardiac output measured by use of lithium dilution

PETCO2

Partial pressure of end-tidal carbon dioxide

PICCO

Cardiac output measured by use of transpulmonary pulse contour analysis

PULSECO

Cardiac output measured by use of the lithium arterial pressure waveform analysis

sCO

Specific cardiac output

a.

Authors. Title (abstr), in Proceedings. 5th Int Cong Vet Anesth 1994;71.

b.

Diazepam, Abbott Laboratories, North Chicago, Ill.

c.

VetaKet, Phoenix Scientific Inc, St Joseph, Mo.

d.

IsoFlo, Abbott Laboratories, North Chicago, Ill.

e.

S/5, Datex-Ohmeda Division, Helsinki, Finland.

f.

Mila International Inc, Florence, Ky.

g.

20-gauge IV catheter, Mila International Inc, Florence, Ky.

h.

Arterial catheter, Pulsion Medical Systems, Munich, Germany.

i.

Baxter Healthcare Corp, Deerfield, Ill.

j.

ABL 660 gas analyzer, Radiometer America Inc, Westlake, Ohio.

k.

LiDCO Ltd, London, England.

l.

Flow-through cell electrode assembly, LiDCO Ltd, London, England.

m.

Peristaltic pump, LiDCO Ltd, London, England.

n.

LiDCO cardiac computer CM 31-01, LiDCO Ltd, London, England.

o.

PICCO, Pulsion Medical Systems, Munich, Germany.

p.

Dobutamine, Bedford Laboratories, Bedford, Ohio.

q.

Medfusion, model 2010i syringe pump, Medex Inc, Duluth, Ga.

References

  • 1.

    Valverde A, Giguère S, Sanchez LC, et al. Effects of dobutamine, norepinephrine, and vasopressin on cardiovascular function in anesthetized neonatal foals with induced hypotension. Am J Vet Res 2006;67:17301737.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 2.

    Valverde A, Giguère S, Morey TE, et al. Comparison of noninvasive cardiac output measured by use of partial carbon dioxide rebreathing or the lithium dilution method in anesthetized foals. Am J Vet Res 2007;68:141147.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 3.

    Tibby SM, Murdoch IA. Monitoring cardiac function in intensive care. Arch Dis Child 2003;88:4652.

  • 4.

    Corley KT, Donaldson LL, Furr MO. Comparison of lithium dilution and thermodilution cardiac output measurements in anaesthetised neonatal foals. Equine Vet J 2002;34:598601.

    • Search Google Scholar
    • Export Citation
  • 5.

    Connors AF Jr, Speroff T, Dawson NV, et al. The effectiveness of right heart catheterization in the initial care of critically ill patients. SUPPORT Investigators. JAMA 1996;276:889897.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6.

    Connors AF Jr, Castele RJ, Farhat NZ, et al. Complications of right heart catheterization. A prospective autopsy study. Chest 1985;88:567572.

  • 7.

    Sandham JD, Hull RD, Brant RF, et al. A randomized, controlled trial of the use of pulmonary-artery catheters in high-risk surgical patients. N Engl J Med 2003;348:514.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8.

    Linton RA, Young LE, Marlin DJ, et al. Cardiac output measured by lithium dilution, thermodilution, and transesophageal Doppler echocardiography in anesthetized horses. Am J Vet Res 2000;61:731737.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9.

    Corley KT, Donaldson LL, Durando MM, et al. Cardiac output technologies with special reference to the horse. J Vet Intern Med 2003;17:262272.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10.

    Giguère S, Bucki E, Adin DB, et al. Cardiac output measurement by partial carbon dioxide rebreathing, 2-dimensional echocardiography, and lithium-dilution method in anesthetized neonatal foals. J Vet Intern Med 2005;19:737743.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11.

    Mason DJ, O'Grady M, Woods JP, et al. Effect of background serum lithium concentrations on the accuracy of lithium dilution cardiac output determination in dogs. Am J Vet Res 2002;63:10481052.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12.

    Beaulieu JM, Sotnikova TD, Yao WD, et al. Lithium antagonizes dopamine-dependent behaviors mediated by an AKT/glycogen synthase kinase 3 signaling cascade. Proc Natl Acad Sci U S A 2004;101:50995104.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13.

    Cooper ES, Muir WW. Continuous cardiac output monitoring via arterial pressure waveform analysis following severe hemorrhagic shock in dogs. Crit Care Med 2007;35:17241729.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14.

    Chen HC, Sinclair MD, Dyson DH, et al. Comparison of arterial pressure waveform analysis with the lithium dilution technique to monitor cardiac output in anesthetized dogs. Am J Vet Res 2005;66:14301436.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15.

    van Lieshout JJ, Wesseling KH. Continuous cardiac output by pulse contour analysis? Br J Anaesth 2001;86:467469.

  • 16.

    Linton NW, Linton RA. Estimation of changes in cardiac output from the arterial blood pressure waveform in the upper limb. Br J Anaesth 2001;86:486496.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17.

    Orme RM, Pigott DW, Mihm FG. Measurement of cardiac output by transpulmonary arterial thermodilution using a long radial artery catheter. A comparison with intermittent pulmonary artery thermodilution. Anaesthesia 2004;59:590594.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18.

    de Wilde RB, Schreuder JJ, van den Berg PC, et al. An evaluation of cardiac output by five arterial pulse contour techniques during cardiac surgery. Anaesthesia 2007;62:760768.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 19.

    Fakler U, Pauli C, Balling G, et al. Cardiac index monitoring by pulse contour analysis and thermodilution after pediatric cardiac surgery. J Thorac Cardiovasc Surg 2007;133:224228.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20.

    Halvorsen PS, Espinoza A, Lundblad R, et al. Agreement between PiCCO pulse-contour analysis, pulmonal artery thermodilution and transthoracic thermodilution during off-pump coronary artery by-pass surgery. Acta Anaesthesiol Scand 2006;50:10501057.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21.

    Jonas MM, Tanser SJ. Lithium dilution measurement of cardiac output and arterial pulse waveform analysis: an indicator dilution calibrated beat-by-beat system for continuous estimation of cardiac output. Curr Opin Crit Care 2002;8:257261.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 22.

    Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res 1999;8:135160.

  • 23.

    Shoemaker WC, Wo CC, Bishop MH, et al. Multicenter trial of a new thoracic electrical bioimpedance device for cardiac output estimation. Crit Care Med 1994;22:19071912.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 24.

    Lin LI-K. A concordance correlation coefficient to evaluate reproducibility. Biometrics 1989;45:255268.

  • 25.

    Lin LI-K. A note on the concordance correlation coefficient. Biometrics 2000;56:324325.

  • 26.

    de Wilde RB, Breukers RB, van den Berg PC, et al. Monitoring cardiac output using the femoral and radial arterial pressure waveform. Anaesthesia 2006;61:743746.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 27.

    Orme RM, Pigott DW. Monitoring cardiac output from the radial artery pressure waveform. Anaesthesia 2007;62:9091.

  • 28.

    Critchley LA, Critchley JA. A meta-analysis of studies using bias and precision statistics to compare cardiac output measurement techniques. J Clin Monit Comput 1999;15:8591.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 29.

    Hallowell GD, Corley KT. Use of lithium dilution and pulse contour analysis cardiac output determination in anaesthetized horses: a clinical evaluation. Vet Anaesth Analg 2005;32:201211.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30.

    Costa MG, Della Rocca G, Chiarandini P, et al. Continuous and intermittent cardiac output measurement in hyperdynamic conditions: pulmonary artery catheter vs. lithium dilution technique. Intensive Care Med 2008;34:257263.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 31.

    Missant C, Rex S, Wouters PF. Accuracy of cardiac output measurements with pulse contour analysis (PulseCO) and Doppler echocardiography during off-pump coronary artery bypass grafting. Eur J Anaesthesiol 2008;25:243248.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32.

    Yamashita K, Nishiyama T, Yokoyama T, et al. Effects of vasodilation on cardiac output measured by PulseCO. J Clin Monit Comput 2007;21:335339.

Contributor Notes

Supported by the State of Florida Pari-Mutuel Wagering Trust and the Florida Thoroughbred Breeders' and Owner Association.

Address correspondence to Dr. Shih.