PulseCO - Pressure Waveform Analysis
History of Pressure Waveform Analysis
History of Arterial Pressure Waveform Analysis
Arterial blood pressure is maintained by ensuring an adequate cardiac output (blood flow) and appropriate vascular resistance (SVR). Mathematically the average / mean arterial pressure MAP = CO*SVR. Cardiac output units of flow are litres per minute. The cardiac output is the product of the stroke volume ejected per beat times the heart rate (beats per minute). The first arterial blood pressure measurement was made in 1733 (by Hale). Continuous arterial blood pressure is now routinely measured in millions of patients per year. Blood pressure monitoring provides a very useful early warning of shock i.e. inadequate blood flow and tissue perfusion. Diagnosing shock and low blood pressure status in surgery allows earlier intervention and resuscitation. Deciding on the correct intervention requires knowledge of the determinants of the low blood pressure. Is the low blood pressure a consequence of heart failure, low blood volume or a change in vascular resistance to blood flow? Concurrently measuring the cardiac output and the arterial blood pressure allows the judgment of a more appropriately choice of corrective intervention. Suitable interventions could include administering fluids - to fill up the vascular volume, giving drugs to support the heart’s strength of contraction, or administration of vasoactive drugs to reduce or increase arterial resistance.
The ease of measuring arterial pressure has meant that the clinical diagnosis and management of the adequacy of the circulation has until this day been very “pressure centric”. Arterial pressure has been used as a surrogate measure of cardiac output. Ideally, diagnostic and clinical decisions concerning the circulation would be based on a full understanding of the drivers of clinically relevant hypo- and hypertensive events observed while monitoring arterial pressure. The goal has been to simultaneously measure cardiac output, blood pressure vascular and systemic resistance.
Early Methods of Deriving Cardiac Output From The Arterial Blood Pressure
Windkessel and “Pulse Contour “Approaches
Not surprisingly the search has been on for a simple method that could provide continuous cardiac output measurements in patients during surgery and in shock. Adolf Fick made the first accurate measurement of cardiac output in 1870. Unfortunately, measuring the cardiac output proved to be not as easy as measuring arterial blood pressure. The “Fick” method proved to be cumbersome and not very practical for most clinical situations.
Otto Frank (1899) was the first person to develop a mathematical model for describing the circulation and the relationship between blood pressure and blood flow. He formulated the concept of the two-element Windkessel model (air chamber/elastic reservoir), which described the shape of the arterial pressure waveform in terms of resistance and compliance elements. His model draws similarities between the heart and the arterial system to a closed hydraulic circuit comprised of a water pump connected to a mixed air and water filled chamber. As water is pumped into the chamber, the water compresses the air in the pocket and simultaneously pushes water out of the chamber, back to the pump. The compression of air simulates the elasticity and extensibility (arterial compliance) of the aorta, as blood is pumped into it from the heart. The resistance that the water overcomes to leave the Windkessel, simulates the resistance to flow encountered by the blood as it flows through the major and minor arteries / arterioles to the capillaries. Functionally, the Windkessel / elastic reservoir effect of the major arteries damps the fluctuation in blood pressure that would otherwise occur if blood was pumped into a stiff and less compliant arterial system. During diastole the stored blood in the aorta and its associated driving pressure, assists in the maintenance of a continuous blood flow away from the heart to the periphery, thereby ensuring consistent end organ perfusion pressure and oxygen transport. This circulation model was improved by Broemser (1930) who proposed a modified 3-element model, which added a second resistive element to the 2-element Windkessel model between the pump and the air-chamber, thereby simulating the resistance to blood flow encountered across the aortic valve.
Frank’s circulatory model and its subsequent refinement led to a number of attempts to derive the stroke volume from more complex Windkessel-like models of the circulation (Kouchoukos, 1970 & Wesseling 1993). These techniques became collectively known as “Pulse Contour” methods of deriving stroke volume and cardiac output from the arterial pressure.
Broemser, Ph., et. al., ``Ueber die Messung des Schlagvolumens des Herzens auf unblutigem Weg'', Zeitung für Biologie 90 (1930) 467-507.
Burattini R, Di Salvia PO (2007) Development of systemic arterial mechanical properties from infancy to adulthood interpreted by four-element Windkessel models. J Appl Physiol 103:66–79.
Erlanger J, Hooker DR. An experimental study of blood-pressure and of pulse-pressure in man. Johns Hopkins Hospital Reports 1904; 12: 145-378.
Fick A. Ueber die Messung des Blutquantums in den Herzventrikeln. Sitzungsberichteder Physiologisch-Medizinosche Gesellschaft zu Wurzburg 1870; 2: 16.
Frank O. (1899) Die Grundform des arteriellen Pulses. Z Biol 37:483–526
Stephen Hales Statical Essays: Haemastaticks, 1733
Langewouters GJ, Wesseling KH, Goedhard WJA (1984) The static elastic properties of 45 human thoracic and 20 abdominal aortas in vitro and the parameters of a new model. J Biomech. 17 (6) 425-435
Langewouters GJ, Wesseling KH, Goedhard WJA (1985) The pressure dependent dynamic elasticity of 35 thoracic and 16 abdominal human aortas in vitro described by a five component model. J Biomech 18:613–620
Remington JW, Noback CR, Hamilton WF, Gold JJ (1948) Volume elasticity characteristics of the human aorta and prediction of the stroke volume from the pressure pulse. Am J Physiology 153: 298-308
Wesseling KH, Jansen JR, Settels JJ et al (1993) Computation of aortic flow from pressure in humans using a nonlinear, three element model. J Appl Physiol 74:2566–2573
The PulseCO algorithm method is based on the principles of conservation of mass and power. Stroke volume is calculated from an analysis of the stroke volume induced pulsatile change in the pressure waveform. It is important to note that this is not a Pulse Contour or modified Windkessel approach, but rather a different and novel non-morphological approach. The PulseCO method overcomes the limitations of the more historic approaches that were reported to be compromised by the variable contribution of the arterial reflected wave and resistance changes that can occur in peripheral arteries. Briefly, the PulseCO analysis transforms the arterial waveform from pressure to a volume equivalent through a compliance and aortic volume correction maneuver. Autocorrelation of the volume waveform derives the beat period (heart rate) and input pulsatile volume change i.e. stroke volume. Cardiac output is derived by multiplying the stroke volume by the heart rate.
The LiDCOplus and the LiDCOrapid monitors both use identical implementations of the PulseCO algorithm. The algorithm derives the pre calibration cardiac output ie CO Algorithm (COa) in exactly the same way in both products.
The COa is then calibrated ie made more accurate by scaling with a calibration factor (CF). This is achieved by entering the CO known (COk) into the monitor. The CF is calculated in both products in exactly the same way CF = COk/COa.
The LiDCOrapid can also derive the CF via a nomogram, which calculates the CF from the patient’s age, height and weight. In this case the displayed cardiac output is then equal to the CF x COa.
Problems addressed that limited the precision of morphology (pulse pressure/systolic area) approaches
Morphology based approaches (Windkessel & Pulse Contour) may have difficulties finding the systolic area
Reflected pressure waves – can move into the systolic part of the waveform with vasoconstriction & further away with vasodilation increasing or
decreasing the area of pressure waveform analysed
Frequent recalibration – after a significant change in hemodynamics is required
Strengths of the PulseCO autocorrelation approach
The power and energy components are conserved within the beat data despite vasoconstriction or dilation, so changes in waveform morphology
& shape do not affect the autocorrelation measurement of stroke volume
Arterial line damping / changes in frequency response do not affect the measurement
Frequent recalibration – after a significant change in hemodynamics - is not required
PulseCO Comparative Studies – Accuracy/Precision
What is the precision/trending ability of the core PulseCO pulse power/autocorrelation algorithm to follow changes in stroke volume & cardiac output?
The software code of the core LiDCO pressure waveform algorithm (PulseCO), as used in the original PulseCO, LiDCOplus and now LiDCOrapid Monitors, has remained completely unchanged since the launch of the first PulseCO monitor in 2001. This means that all the 100 or so papers and abstracts published on the performance of the core PulseCO software are comparable and still relevant.
Acceptable limits of precision have been defined by Critchley & Critchley (1999), who stated that if a new (cardiac output) method is to replace an older, established method, the new method should have errors not greater than the older reference method. Over the last 10 years the precision of LiDCO’s core ‘PulseCO’ algorithm to trend changes in stroke volume has been evaluated in a wide number of challenging clinical situations – these include: general surgical patients (Heller et al., 2002), high cardiac outputs (Hallowell & Corley 2005), hyperdynamic liver transplantation patients (Costa et al., 2007), off-pump (Missant and Wouters 2007) and on-pump cardiac surgery (Wilde et al., 2007; ), post pediatric heart transplantation (Kim et al., 2006), post-operative care (Pitman et al., 2005, Hamilton, Huber and Jessen, 2002); pre-eclampsia (Dyer et al., 2011), congestive heart failure (Kemps et al., 2008, Mora et al 2011) and general intensive care (Mills et al ., 2010, Brass et al., 2011, Cecconi et al., 2010, Smith et al., 2005).
As can be seen (Table below) the 95% confidence limits (Bland Altman statistics) reported in these studies range between 17% – 30% in adults and are within the original precision limits proposed by Critchley and Critchley (1999). Therefore, these results obtained from a variety of clinical situations demonstrate that the PulseCO algorithm has sufficient precision to follow cardiac output changes without recalibration and is equivalent in precision to bolus thermodilution.
Summary of key studies comparing PulseCO algorithm to other validated cardiac output measurements Differences to control technique Limits of agreement Study & Control No. of patients No. Obs. Bias L/min Precision L/min 95% c.l. L/min & % Lower L/min Upper L/min (1) De Wilde et al 2007
(a) COtd, intra op cardiac
(b) Literature survey of
PulseCO pub. data
(2) Missant, et al 2007
COtd intra op. off-pump
20 149 -0.03 0.65 1.3
-1.33 1.26 (3) Costa et al 2007
COtd – post op 48 hrs
hyperdynamic liver trans.
23 151 0.29 1.09 2.17
-1.87 2.46 (4) Kirwan et al 2005
(a) Lithium dilution comp.
(b) PiCCO vs PulseCO 24
(5) Pittman et al 2005
Lithium dilution comp.
ICU 24 hrs post surg
21 83 -0.01 0.82 1.64
-1.65 1.63 (6) Hamilton et al 2002
Post cardiac surg. (8hrs)
20 80 0.1 0.6 1.2
-1.1 1.3 (7) Kemps et al., 2008
Exercising congestive heart
failure - patients
10 17 - 26 Rest -1%
N/A Rest -27.5 %
(9) Dyer et al 2010
18 N/A 0.16 0.68 1.35
The data shown above in the table are derived from the statistical analysis of repeated paired measures at differing cardiac outputs from multiple patients. The statistics used assumes each pair of measurements are fully independent from each other (conventional method). In fact, they are not fully independent as they are actually multiple measures taken within a single patient. Thus the conventional method of analysis (Bland Altman/Critchley) is by far the most commonly used way of comparing the accuracy of two methods of measurement of cardiac output. Another, and perhaps more clinically relevant way of looking at the data is to examine the way that changes in cardiac output from each of the sequential PulseCO measurements compares to the changes in sequential dilution cardiac output control measurements. This is known as the consecutive change method of analysis. This form of analysis was used in the paper published by Wilde et al., 2007. Some of their results are shown below. The advantage of this consecutive change analysis method is that it can be used to quantify the concordance (another way of looking at the trending ability) between the two techniques. It can be seen that 88% of the relative change comparison points are seen to be falling in the upper right and lower left quadrants of the plot. This analysis allowed Wilde et al., to state that 88% of all changes in cardiac output of greater than 0.5l/min were similarly detected by the two methods i.e. PulseCO and the control thermodilution measurements.
Comparison against Electromagnetic flow probe & Fick
1. Marquez J, McCurry K, Severyn D, Pinsky M. (2008). Ability of Pulse Power, Esophageal Doppler and Arterial Pulse Pressure to Estimate Rapid Changes in Stroke Volume in Humans. Crit Care Med 36(11) 3001 – 3007
2. Kemps H, Thijssen E, Schep G, Sleutjes B,De Vries W, Hoogeveen A, Wijn P, Doevendans P. (2008) Evaluation of two methods for continuous cardiac output assessment during exercise in chronic heart failure patients. J Appl Physiol 105: 1822-1829
Comparison against pulmonary artery thermodilution
3. Hamilton TT, Huber LM, Jessen ME. (2002) PulseCO: A Less-Invasive Method to Monitor Cardiac Output From Arterial Pressure After Cardiac Surgery. Ann Thorac Surg; 74:S1408-12
4. Pittman J, Bar Yosef S, SumPing J, Sherwood M, Mark J (2005) Continuous cardiac output monitoring with pulse contour analysis: A comparison with lithium indicator dilution cardiac output measurement. Crit Care Med. 33 (9) 2015-2021
5. Missant C, Rex S, Wouters P. (2008) Accuracy of cardiac output measurements with pulse contour analysis (PulseCO) and Doppler echocardiography during off-pump coronary artery bypass grafting. European Journal of Anaesthesiology 25( 3): 243-248
6. Costa MG, Della Rocca G, Chiarandini P, Mattelig S, Pompei L, Barriga MS, Reynolds T, Cecconi M, Pietropaoli P (2007) Continuous and intermittent cardiac output measurements in hyperdynamic conditions: pulmonary artery catheter versus lithium dilution technique. Intensive Care Medicine DOI 10.1007/s00134-007-0878-6
7. Wyffels P, Sergeant P, Wouters P (2010) The value of pulse pressure and stroke volume variation as predictors of fluid responsiveness during open chest surgery. Anaesthesia 65: 704:709j
8. Dyer R, Piercy J, Reed A, Strathie G, Lombard C, Anthony J, James M (2011) Comparison between pulse waveform analysis and thermodilution cardiac output determination in patients with severe pre-eclampsia. Brit Journal of Anaesthesia 106 (1) 77 – 81
9. De Wilde RBP, Schreuder JJ, van den Berg PCM, Jansen JRC (2007) An evaluation of cardiac output by five arterial pulse contour techniques during cardiac surgery. Anaesthesia: 62, 760-768
10. Kirwan C, Smith J, Lei K, Beale R (2005) A comparison of two calibrated continuous arterial pressure waveform based measurements of cardiac output over 24 hour. Crit Care Med. 33(12) Suppl; 208-S; A56