Chapter 21 Principles of Invasive Monitoring

Jeremy S. Hertzig, Jacqueline M. Evans, Kenneth A. Schenkman

• Hemodynamic monitoring refers to measurement of the functional characteristics of the heart and the circulatory system that affect the perfusion of tissues with oxygenated blood.

• Hemodynamic monitoring can be performed invasively or noninvasively and can be used for diagnosis, surveillance, or titration of therapy.

• The central venous waveform is composed of three waves (a, c, and v) and two wave descents (x and y).

• The arterial waveform has three components: rapid upstroke, dicrotic notch, and runoff.

• Cardiac output can be calculated using the Fick method or measured directly via thermodilution.

• Continuous SvO₂ monitors can guide goal-directed therapy in patients who are in septic shock.

• A pulmonary artery catheter can be used to measure cardiac output and indices of oxygen delivery and extraction.

Role of Invasive Hemodynamic Monitoring

Since William Harvey’s observation in the early 1600s that the heart pumps blood in a continuous circuit, the function of the circulatory system has been the subject of intense scrutiny. Hemodynamic monitoring refers to measurement of the functional characteristics of the heart and the circulatory system that affect the perfusion of tissues with oxygenated blood in order to maintain homeostasis and to remove byproducts of metabolism. Several different types of invasive hemodynamic monitoring can be used concurrently to guide management. The goal of hemodynamic monitoring is to provide more accurate diagnoses and to guide additional interventions to deliver improved care to the critically ill patient.

In his 1733 report “Statical essays: containing haemastaticks; or, an account of some hydraulick and hydrostatical experiments made on the blood and blood-vessels of animals,” Hales ¹ described early experiments in horses in which he was the first to measure central venous pressure (CVP). Figure 21-1 depicts Hales and an assistant in the process of these early experiments.

Figure 21–1 Clinician and an assistant measuring the blood pressure of a horse.

(From Pickering G: Systemic arterial hypertension. In Fishman AP, Dickinson WR, editors: Circulation of the blood: men and ideas, New York, 1964, Oxford University Press.)

Clinical hemodynamic assessment at the bedside begins with noninvasive measurements such as heart rate (HR), blood pressure, urine output, and peripheral perfusion. Other noninvasive studies that may contribute to assessment of hemodynamic status include electrocardiograms, chest radiographs, and echocardiography. Frequently, in the pediatric intensive care unit (ICU) these measurements are supplemented by invasive hemodynamic measures that require entrance into the intravascular space. Such invasive hemodynamic measurements include placement of central venous catheters to assess right atrial filling pressures and to measure mixed venous oxygen saturation, arterial catheters to assess arterial blood pressure, and pulmonary artery catheters to assess left-sided pressures, cardiac output (CO), and vascular resistance. Although invasive hemodynamic monitoring can provide the skilled intensivist with a plethora of valuable information, it is not meant to take the place of or minimize the extensive amount of information that can be gained by less invasive techniques. Successful use of invasive hemodynamic measurements necessitates the requisite skills to obtain these measures safely and with utmost attention to the multiple potential risks imposed upon the patient. Furthermore, for invasive hemodynamic measurements to be useful, the clinician must be able to successfully interpret the information provided by the measurements. Finally, as with any technology, the use of invasive hemodynamic monitoring is in evolution, and it is incumbent upon the clinician to be familiar with developments as they arise and with current controversies regarding these procedures.

This chapter aims to be a practical guide to the use of hemodynamic monitoring in the pediatric ICU. The chapter reviews general principles of measurement and then discusses the three main types of invasive hemodynamic monitoring: central venous catheter, arterial catheter, and pulmonary artery catheter. It addresses the indications and controversies, sites of insertion, interpretation of waveforms, and potential complications. It also reviews CO monitoring and calculation of oxygen consumption and delivery. Chapter 15 discusses the specific techniques for gaining access in order to make these measurements.

Indications for Invasive Hemodynamic Measurements

The three main indications for invasive hemodynamic monitoring are diagnosis, surveillance, and titration of therapy. Diagnosis may include the differentiation of septic shock (through assessment of factors such as diminished right heart filling pressures or preload and decreased systemic vascular resistance) from cardiogenic shock (which is characterized by elevated left heart pressures and afterload). Surveillance implies observation over time. The purpose of surveillance may be to assess the stability of a patient at risk for adverse changes or to determine the response to therapy. Invasive measurements performed for diagnostic purposes often are continued for surveillance purposes. Titration of therapy often is based on information gleaned from invasive measurements.

Principles of Measurement

Intensive care clinicians rely on a wide variety of measurement systems to assess patient clinical status and response to therapy. However, not all clinicians have a good understanding of how physiologic variables are measured, and some may not be able to troubleshoot monitoring systems or recognize when information obtained is inaccurate. A detailed discussion of monitoring is beyond the scope of this chapter, but a basic understanding of the principles of measurement is helpful in deciding which measurements to trust and how to assess a monitoring system for accuracy. Detailed descriptions of monitoring systems are provided elsewhere.²^–⁴

Signal Analysis

Measurements generally are made directly by comparison with known standards or indirectly by use of a calibration system. Determination of length or weight usually is made by direct comparison with a standard ruler or standard mass. Most invasive measurements in the ICU are made indirectly, therefore requiring use of a calibration system. Thus understanding the basis for calibration of a system is important to determine the validity of the measurement.

Measurement systems detect and transform signals so that they can be presented in an interpretable way to the user. Signals can be characterized as static or dynamic. Slowly changing signals, such as body temperature, can be thought of as static. Hemodynamic measurements change from moment to moment and thus are dynamic. Physiologic signals may be periodic; for example, arterial pressure is periodic because it varies with the cardiac cycle.

Complex periodic signals, such as an arterial pressure waveform, can be described mathematically as the sum of a series of simpler waveforms called a Fourier series. Alternatively, the arterial tracing can be thought of as a sum of simpler waveforms, sine waves, and cosine waves. Figure 21-2 depicts an arterial pressure waveform as the sum of the first six terms in the Fourier series. The sum of the first six terms in the series forms a waveform similar to the original tracing. Adding additional terms from the Fourier series, or higher harmonics, results in an increasingly better representation of the actual waveform. In general, to reproduce a pressure tracing without loss of significant characteristics for clinical use, the measurement system must have an accurate frequency response to approximately 10 times the fundamental frequency (first 10 harmonics).

Figure 21–2 Fourier series representation of an arterial pressure tracing. Bottom, High-fidelity carotid artery pressure tracing and the sum of the first six harmonics of its Fourier series representation. Despite the few terms used in the synthesis, the close fit of the two curves is evident. Top, Individual harmonic components labeled with their harmonic number.

(Redrawn from RSC: Transducers for biomedical measurements: principles and applications. New York, 1974, John Wiley & Sons, from Hansen AT: Acta Physiol Scand 19[suppl 68]:1, 1949; and Perloff WH: Invasive measurements in the PICU. In Fuhrman BP, Zimmerman JJ, editors: Pediatric critical care, ed 2, St. Louis, 1998, Mosby.)

The sampling rate of a measurement system determines how often a physiologic value is measured. For body temperature, sampling every few minutes might be sufficient, but for arterial pressure measurement, a higher rate is needed. This principle may seem obvious, but as an example of the importance of sampling rate, consider the number of points needed to define a circle. If we place three equidistant points on a circle, we describe a triangle, not a circle. Similarly, four points describe a square. If we increase the number of points (sampling rate), we can describe the circle more completely. For a sine wave, the minimum frequency of sampling needed to preserve the waveform is twice the frequency. This mathematical minimum is known as the Nyquist frequency.² For complex waveforms such as arterial pressure tracings, the sampling rate must be at least twice the highest frequency component in the waveform.

Measurement Systems

Hemodynamic monitoring in the clinical setting usually uses a fluid-coupled system where changes in pressure are transmitted via a column of (uncompressible) fluid in a (ideally incompressible) tube to a mechanical transducer. The mechanical transducer, usually a displaceable screen diaphragm, converts a change in pressure to an electrical signal, which can be processed and displayed. In laboratory settings, vascular pressures can be measured by a transducer at the point of interest rather than remotely as in the clinical setting. Measuring pressure at the point of interest—directly in the aorta, for example—decreases loss of signal integrity because of the measurement system. Most clinical pressure measuring systems have sufficient fidelity for clinical purposes. However, compliance, resistance, or impedance in the pressure tubing can result in damping or alteration of the recorded signal. The presence of bubbles in the fluid can further damp the recorded signal.

Errors in Measurement

The ideal measurement system determines the actual or “true” value for the measured variable. However, determination of a true value may be difficult. Every measurement system is subject to various errors. Errors in measurement can be classified as either systematic or random. Systematic errors occur in a predictable manner and are reproduced with repeated measures. Bias in a measurement system—for example, a baseline offset—results in a systematic error. Random errors are unpredictable and do not recur predictably with repeated measures.

Accuracy of a measurement is defined by the difference between the measured and true values, divided by the true value. Precision is defined by the reproducibility of the measurement; thus a more precise system yields more similar values for repeated measures under the same conditions than does a less precise system. Imprecision can be thought of as a representation of random errors, whereas bias can be thought of as a representation of systematic errors.

Calibration

Many measurement systems are linear, that is, based upon an assumption that the relationship between the inputs and outputs from a measurement device can be fitted to a straight line. This assumption allows a system to be calibrated under two conditions, with the rest of the values falling on the line defined by those two points. Actual nonlinearity of the system adversely affects the measurements.

Calibration is a process in which the reading, or output of a device, is adjusted to match a known input value. For example, an electronic pressure transducer may be calibrated against a mercury manometer. If the input to the device is zero, the output should be adjusted so that the reading also is set to zero. This “zeroing” reduces any baseline offset, thus reducing systematic errors in subsequent readings. The system then is calibrated to a nonzero value, for example, 100 mm Hg pressure, and the system gain is adjusted to read this value as well.

Frequency Response

The ability of a measurement system to accurately measure an oscillating signal, such as arterial blood pressure, is dependent upon the system’s frequency response. The system can either overestimate or underestimate the true amplitude of a signal. If the system is overdamped, the value reported underestimates the amplitude, and waveform characteristics may be lost. Resonance in the system may result in overestimation of the amplitude. Measurement of arterial systolic pressure—the amplitude of the arterial waveform—may be inaccurate because of overdamping, and important waveform characteristics may be lost if the frequency response of the measurement system is poor. Figure 21-3 illustrates the effects of dampening on measurement of blood pressure.

Figure 21–3 Effects of dampening on blood pressure measurement. Upper two graphs depict the response to a square-wave input from three different blood pressure transducers with different dampening. Lower two graphs show the effect of dampening on blood pressure measurement.

(From Chatburn RL: Principles of measurement. In Tobin MJ, editor: Principles and practice of intensive care monitoring, New York, 1998, McGraw-Hill.)

Impedance

Impedance is the ratio of the change in blood flow along a vessel to the change in the pressure in the vessel. Impedance has both resistive and reactive components. In a pulsatile system such as the cardiovascular system, resistance alone does not fully describe the impediment or impedance to forward flow of blood. The caliber, length, and arrangement of the blood vessels and the mechanical properties of the blood determine resistance in the blood vessels. Reactance includes compliance of the vessels and inertia of the blood and thus is a dynamic component of impedance. This is important because the pulsatile nature of the cardiovascular system is dynamic.

When blood is propelled through a vessel at a branch point, a reflected pressure wave back toward the heart increases the impedance of the system. The major sites of wave reflection from vessel branching are from vessels approximately 1 mm in diameter.³ Thus these small vessels contribute significantly to overall impedance. Figure 21-4 shows the relationships between pressure and flow velocity with distance along the length of the aorta. Because blood pressure increases with distance from the heart and flow velocity decreases with distance, the impedance increases toward the peripheral vasculature. Hemodynamic measuring systems are essentially physical extensions of the vascular system; thus the configuration and characteristics of the tubing and transducer system can alter the overall effect of impedance.

Figure 21–4 Pressure pulses and flow velocity at various points in the systemic arterial circulation. Data were obtained from dogs and are similar to measurements made in humans. The data indicate that both peak and pulse pressure increase with distance from the heart, whereas oscillation in flow velocity shows a progressive decrease. Consequently impedance (discussed in the text) must increase toward the periphery.

(From Perloff WH: Invasive measurements in the PICU. In Fuhrman BP, Zimmerman JJ, editors: Pediatric critical care, ed 2, St Louis, 1998, Mosby.)

Invasive Techniques

Central Venous Pressure Catheters

Indications

Indications for CVP catheter placement in pediatric patients include assessment of right heart filling pressure (CVP), monitoring of large fluid shifts between the intravascular and the extravascular space, infusion of vasoactive substances, monitoring mixed venous oxygen saturation for goal directed therapy of sepsis, and infusion of hyperosmolar fluids and/or irritants.⁵^–⁷

Interpretation of Waveforms

CVP is a measure of right atrial pressure, although it may be measured in the inferior or superior vena cava (SVC). It is a measure of preload—the force or load on the right ventricle during relaxation or filling. CVP is measured at the end of diastole, just prior to ejection. Final filling of the right ventricle occurs at the end of atrial contraction. When the tricuspid valve is open during diastole, the right atrium and right ventricle form a continuous column; therefore right atrial pressure reflects right ventricular end-diastolic pressure. CVP is used to measure filling pressure or preload and as such is an indicator of volume status. It is commonly used in patients with hypovolemic or septic shock in whom volume resuscitation is desirable prior to institution of vasopressor therapy. In patients with decreased right ventricular function or pulmonary hypertension, an increased CVP well beyond normal limits may be observed, and further fluid resuscitation may promote the development of congestive heart failure. Increases in the positive end-expiratory pressure can decrease preload despite a paradoxically increased CVP. Finally, increases in extrathoracic pressure, such as that caused by increased abdominal distension, can increase CVP.

The CVP waveform is divided into three components: a, c, and v waves (Figure 21-5). Each component can be correlated with a specific portion of the electrocardiogram (ECG) tracing. The a wave occurs with atrial contraction and is seen after the P wave of the electrocardiogram during the PR interval. Thus the mean value of the a wave approximates right ventricular end-diastolic pressure. Canon a waves (Figure 21-6), which are enlarged a waves seen when the right atrium is ejecting against a closed tricuspid valve, may be seen when atrioventricular discordance occurs (i.e., during junctional ectopic tachycardia, ventricular tachycardia, or heart block). The c wave occurs in early systole with closure of the tricuspid valve and is seen at the end of the QRS complex in the RST junction. The v wave occurs during filling of the right atrium in late systole prior to opening of the tricuspid valve and is seen between the T and P waves of the ECG. The v wave is increased in the setting of tricuspid regurgitation. The x descent is the decrease in pressure after the a wave, reflecting atrial relaxation. The y descent is the decrease in pressure that occurs after the v wave as the tricuspid valve opens and passive filling of the right ventricle occurs.

Figure 21–5 Central venous pressure (A) tracing with corresponding electrocardiogram (ECG). The a wave is produced by atrial contraction and occurs after the P wave of the ECG during the PR interval. The c wave (C) is produced by closure of the tricuspid valve and takes place early in systole at the end of the QRS complex in the RST junction. The v wave (V) is caused by rapid filling of the right atrium late in systole prior to opening of the tricuspid valve and is seen between the T and P waves of the ECG. The x descent (X) reflects the decrease in pressure in the right atrium after the a wave as the tricuspid valve is pulled away from the right atrium by the right ventricle as it contracts during systole. The y descent (Y) is the decrease in right atrial pressure that occurs after the v wave as the tricuspid valve opens and blood moves from the right atrium into the right ventricle.

(From O’Rourke RA: The measurement of systemic blood pressure: normal and abnormal pulsations of the arteries and veins. In Hurst JW, editor: The heart, New York, 1990, McGraw-Hill.)

Figure 21–6 Canon a waves are enlarged a waves seen when the right atrium is ejecting against a closed tricuspid valve. These waves are typically seen when atrioventricular discordance occurs, such as during junctional ectopic or ventricular tachycardia or heart block.

Continuous Mixed Venous Oxygen Saturation

Mixed venous oxygen saturation (SvO₂) can be measured continuously by using a specially designed central venous catheter. These catheters have two to three lumens and have the same capabilities of the catheters previously described. The SvO₂ catheters use reflection spectrophotometry and are able to read hemoglobin oxygen saturation continuously. The reflected light is dependent on the oxygenated and deoxygenated hemoglobin concentration in the circulating blood.⁸

SvO₂ is another parameter used to monitor the relationship between oxygen delivery and demand and is often used as a surrogate for cardiac index. Rivers et al.⁹ showed that when continuous SvO₂ monitoring was used to guide resuscitation and hemodynamic support in patients with severe sepsis and septic shock, survival rates improved. Recent guidelines set forth by the American College of Critical Care Medicine/Pediatric Advanced Life Support have recommended goal-directed therapy with a target SvO₂ of ≥70% in children and adolescents who are in septic shock.⁷ A randomized controlled trial conducted by Oliveira and colleagues¹⁰ supported the use of these guidelines in children and adolescents with severe sepsis or fluid-refractory septic shock.

Ideally, the catheter should be placed in the right internal jugular, with the tip taking measurements in the SVC. SvO₂ measurements obtained from the inferior vena cava exhibit greater variability because of fluctuations in splanchnic oxygen utilization and thus are less reliable. SvO₂ measurements from the right atrium contain coronary sinus blood and are more desaturated because of the high oxygen extraction rate of the myocardium. Studies in critically ill children have evaluated SvO₂ measurements obtained in the pulmonary artery and the SVC. Concordance analysis showed appropriate agreement in the measurements between these two sampling sites.¹¹ This finding has clinical importance because the use of pulmonary artery catheters (PACs) has declined, while central venous line use has increased.¹² Continuous SvO₂ monitoring can alert the intensivist to early changes in hemodynamic status and allows for less frequent opening of the central line for blood sampling and thus less risk of infection. Percutaneously placed SvO₂ central lines have even been used to monitor patients undergoing complex cardiac surgery, thus avoiding the risks associated with transthoracic lines following surgery.¹³

Arterial Pressure Catheters

Indications

The transition to direct monitoring of arterial blood pressure dates back to the mid 1950s when two separate studies compared invasive arterial measurements and noninvasive or cuff measurements in healthy adults.^14,¹⁵ Van Bergen and colleagues¹⁵ noted a frequent difference between direct and indirect measurements, with indirect measurements increasingly lower than direct measurements as the systemic blood pressure increased. The greatest disparity was found in young hypertensive patients. Similarly, Cohn and Luria¹⁶ observed that invasive arterial pressures were significantly greater than cuff pressures and emphasized the importance of direct measurements of systemic arterial pressure when caring for patients with hypotension and shock. Continuous direct monitoring of arterial blood pressure should be considered when treating patients who require more than minimal vasopressor therapy.

Indications for arterial catheterization include continuous monitoring of systemic arterial blood pressure, frequent blood sampling, and withdrawal of blood during exchange transfusions.¹⁷ In addition to the value of the measurements themselves, these measurements provide components of derived measures of CO and oxygen delivery.

Interpretation of Waveforms

Systolic blood pressure (SBP) in children varies greatly with age and gender. As with the CVP waveform, the arterial waveform can be correlated with specific parts of the cardiac cycle. The arterial waveform has three main components (Figure 21-7): (1) a rapid upstroke and downslope that correlates with systolic ejection, (2) a dicrotic notch that correlates with closure of the aortic valve, and (3) a smooth runoff that correlates with diastole. The dicrotic notch or incisura is decreased in situations of hyperdynamic CO in which left ventricular output and stroke volume (SV) are increased, pulse pressure is widened, and diastolic blood pressure (DBP) is increased (e.g., surgical systemic-to-pulmonary shunts, patent ductus arteriosus, aortic regurgitation, anemia, fever, sepsis, hypovolemia, and exercise).¹⁸ Conversely, cardiac tamponade and severe aortic stenosis can narrow the pulse pressure and are associated with a deflection (anacrotic notch) on the ascending limb of the waveform.¹⁸