Clinical measurement and monitoring

The ability to measure and monitor patients’ physiology is fundamental to modern anaesthetic practice, and a variety of sophisticated instruments are available. It is crucial that the anaesthetist understands not only the data being generated but also the limitations of any equipment and potential sources of error. Furthermore, the anaesthetist must have the knowledge and experience to integrate multiple clinical measurements. A comprehensive understanding of each monitoring system is therefore crucial to ensure optimal patient care and avoid potentially harmful mistakes.

Clinical measurement is limited by four major constraints:

Feasibility. It is often technically difficult to record a physiological variable (e.g. stroke volume) reliably or accurately.
Reliability. This depends on the device being calibrated and used correctly (e.g. proper placement of ECG electrodes, appropriate sizing of the NIBP cuff). Delicate equipment, such as a blood gas analyser, requires regular maintenance and calibration.
Interpretation. Any measurement must be understood correctly as part of a complex physiological system. Arterial pressure may be within the normal range despite severe hypovolaemia; global measurements of end-tidal carbon dioxide tension or oxygen saturation are influenced by many factors other than ventilation. Anaesthetists must integrate all this information when assessing a patient's physiology.
Value in improving patient care. This includes the ease, convenience and usefulness of a measurement and evidence of improvement in patient safety and outcome.

Monitoring represents the assessment and use of measurements to direct therapy. Monitors usually comprise four components ( Box 17.1 ):

A device that connects to the patient
A measuring device (often a transducer that converts a measurement into electrical signals)
A computer, which amplifies and filters the signal, then integrates it with other variables to produce useful information
A display showing the results as a wave, number or combination

Box 17.1

Four components of a monitor

Connection to patient

Measuring device

Electronic filter/amplifier

Display

Note that monitors often do not directly measure the displayed variable and that the displayed variable may not reflect the underlying physiology. For example, an ECG does not measure cardiac function; therefore a normal ECG trace does not indicate that the heart is pumping effectively. When interpreting measurements the following questions should be asked:

What is being measured? Whilst arterial pressure is a direct measurement, ‘depth of anaesthesia’ is less obvious. Furthermore, a monitor may display a variable (e.g. heart rate) using different sources (e.g. pulse oximeter, ECG, arterial waveform) and so can change in the absence of any physiological derangement.
How is it measured? Arterial pressure is often measured by either a transducer attached to an arterial cannula or an automated oscillometer. The values between each can differ, and this difference should be accounted for when monitoring the patient.
Is the environment appropriate? Many monitors are designed for use in operating theatres and may not function correctly if exposed to cold and vibration (e.g. in an ambulance or helicopter). The magnetic field of an MRI scanner is also a hostile environment.
Is the patient appropriate? Monitors designed for adults may be inaccurate when used on small children. Obese adults may require a large blood pressure cuff, and ECG readings may be of low quality.
Has the monitor been applied to the correct part of the patient? For example, in aortic coarctation, arterial pressure may be markedly different in each arm. Pulse oximeters also fail to operate reliably if placed distal to a blood pressure cuff.
Is the variable within the range of the monitor? Most monitors are validated on healthy patients in laboratories. They are not necessarily accurate at physiological extremes such as in anaphylaxis or septic shock or outside the limits of a healthy individual (e.g. low saturations in pulse oximetry).
Has the monitor been checked, serviced and calibrated at the correct intervals? Regular servicing and calibration are expensive and time consuming. All equipment should be tagged with a service sticker that identifies the service date, when the next service is due and who to contact in case of malfunction. Equipment should not be used if it has not been serviced or is past its service date.

Box 17.2 shows the checks that the anaesthetist should follow before using a patient monitor.

Box 17.2

Premonitoring checks

What is being measured?

What method is being used?

Has the monitor been serviced and calibrated?

Is the environment appropriate?

Is the patient appropriate?

Is it attached to the appropriate part of the patient?

Is the range appropriate?

Can the display be read?

Are the alarms on and have the limits been set?

Process of clinical measurement

Stages of clinical measurement

There are four stages of clinical measurement:

Detection of the biological signal by a sensor
Transduction, which is conversion from one form of energy (the sensor output) to another (usually electrical)
Amplification and signal processing to extract and magnify the signal and reduce unwanted noise
Display and storage of the output; whilst commonly the electronic representation of a biological signal, this also includes the height of a fluid-column manometer for pressure measurement, expansion of alcohol in a thin glass column for temperature measurement, or a mechanical recording for peak-flow measurements

Essential requirements for clinical measurement

Clinical measurement devices detect a biological signal and reproduce it in a convenient display or recording. The conversion of a biological signal into an electrical recording introduces some key concepts: linearity, drift, hysteresis, signal-to-noise ratio, static and dynamic response, and accuracy and precision.

Linearity describes the response of a measurement system to changes in the biological input signal; for example, if the true mean arterial pressure of a patient increases by a factor of 0.5 (MAP from 60–90 mmHg) the measurement device will respond by a factor of 0.5. In practice there is a limited range in which the relationship is linear; this is usually provided by the manufacturer.

Hysteresis occurs where the response of the measurement device has some dependence on previous measurements, thereby altering linearity. For example, temperature measurement can be affected in some systems by whether the patient's temperature is increasing or decreasing, resulting in measurements that under or overread the true temperature. Bimetallic strips used for temperature compensation in vaporisers are particularly prone to hysteresis; the rate of deformation of the strip on heating is different to the rate of deformation when the strip then starts to cool.

Drift describes the slow increase or decrease of measurement values when there is no change in the underlying biological value. Pressure transducers often show drift, with the zero-point changing over time, caused by heating of the electrical components affecting their resistance. Regular calibration in the form of setting the zero point against atmospheric pressure minimises the clinical impact of drift. Two-point calibration – calibrating against zero and another point at the maximum value in the device's operating range – further protects against drift affecting the linearity of response. More points can be used to ensure accuracy ( Fig. 17.1 ).

Fig. 17.1, A linear response is seen in A where a change in the true value is matched by a proportional change in the measured value. B demonstrates zero drift, with B1 being a proportional change to the whole linear response, and B2 demonstrating drift affecting zero more than the higher values. A 2-point or greater calibration protects against B2 . Hysteresis is seen in C ; true values are underread non-linearly in the ascent, with true values being overread in the descent. This pattern is seen with bimetallic strips in temperature measurement.

Signal-to-noise ratio (SNR) reflects the degree by which the measured signal is affected by other patient or environmental signals. For example, the EEG measures tiny neuronal voltages and can easily be overwhelmed by larger voltages from the patient (muscle electrical activity) or environment (mains electricity). Increasing the SNR using filters, signal processing and isolation devices ensures that the measured EEG potentials accurately reflect the biological neuronal electrical signals.

Response can be either dynamic or static. Most medical devices display a dynamic response, which is a changing measurement to a rapidly changing underlying signal, as seen with arterial pressure transduction. A static response is commonly seen for single values or values that change very slowly over time, such as temperature. Importantly, in static systems there is time for the measurement device to settle around the true value, unlike dynamic systems where a slow response can lead to inaccurate measurements.

Accuracy is the difference between a measurement and the actual physiological variable, usually determined by a gold standard measurement. Calibration enhances accuracy, usually tested against a known value, such as the zero-reference point for direct arterial pressure measurement.

Precision is the reproducibility of repeated measurements. Single recordings are unreliable for imprecise measurements, particularly when a test requires patient cooperation, practice or effort, such as peak expiratory flow rate. Repeated measurements demonstrate the variability in response ( Fig. 17.2 ).

Fig. 17.2, (A) Accuracy is represented by this target; the dots group round a centre (true) point, but not to each other. (B) Precision is represented by this target; the dots are grouped together but not on the centre point.

There are no perfect monitors. Even if the reported value is ‘true’, it must be interpreted in the context of the patient and their environment. Treat the patient, not the monitor is an old adage that remains true.

The importance of repeated measurements

Differences in repeated clinical measurements arise from three causes:

Change within the patient
Inherent variability in the signal or its measurement
Confounding errors – the recorded measurement does not reflect the signal

Consistent repeated measurements ensure precision but not accuracy (e.g. invasive blood pressure recordings may be consistent but inaccurate if the transducer is not correctly calibrated). In general calibrated instruments are accurate but not necessarily precise (e.g. cardiac output monitors are generally either accurate but imprecise if they require calibration or inaccurate but precise if they do not). Therefore anaesthetists should be aware of the accuracy and precision of measurement systems, and when a measurement does not fit the clinical picture, use a different technique to confirm or refute the initial reading.

Measurement of continuous signals over time

Continuous signals (as used in most modern measurement devices) also need an assessment of the response of the system to a changing biological signal. The reliability of such signals is determined by their input–output relationship. Accuracy requires good zero and gain stability, a linear amplitude response and minimal hysteresis, and an adequate frequency response. These concepts will be explored later, but it is important to realise environmental changes can influence the response of a measurement device to a changing biological signal (e.g. humidity may affect capnography measurements).

Analogue and digital measurement

Modern measurements usually transduce an analogue biological signal to an electrical signal, but mechanical devices still have an important role. For example, peak flow meters, which are convenient and need no electrical source, use the physics of gas flow to move a calibrated indicator. Analogue computers, comprising electronic circuits and operational amplifiers, are also still in use, primarily in the electronics industry in the form of oscilloscopes. However, modern high-powered digital computers have overtaken most of the roles previously undertaken by analogue systems.

Digital signal processing

Digitisation is the process of converting a continuous analogue signal into discrete values, achieved by sampling the signal at regular intervals. A waveform can be broken up into a series of separate points; the frequency of sampling determines how close together these points are and represents the resolution of the digital device (see later). The advantage of digital processing is that the digitised signal can be manipulated and analysed using sophisticated software calculations.

Analogue-to-digital conversion

Computers process information in discrete form, operating on data expressed as either 0 or 1 (i.e. binary code). The resolving power of a computer is limited by the maximum number of digits that can be represented in binary code. For example, an 8-bit computer processor can use binary code to represent 2 ⁸ decimal numbers from 00000000 (or 0 as a decimal number) to 11111111 (equivalent to 255 as a decimal number). An analogue signal can be resolved by an 8-bit converter with an accuracy of 1 part in 255, or 0.4%. More modern processors capable of 32-bit conversion therefore can represent 2 ³² decimal numbers, giving a resolving power of 0.00000002%. In the domestic market 64-bit computing is now common. This accuracy comes at a cost in terms of hardware, power consumption and storage requirements.

Sampling frequency is an important part of signal resolution. A low sampling frequency may be adequate for a slowly changing waveform, but it may not be representative of high-frequency components. This introduces an aliasing error in which different signals become indistinguishable. According to the Nyquist theorem, the sampling frequency should be at least twice the component of the input signal waveform with the highest frequency and sufficient amplitude; for example, a sampling frequency of 100 Hz would adequately capture the fastest rate of change in a physiological pressure signal.

Data display

Data can be presented in either analogue or digital form. A type of analogue display is a mechanical spirometer that records flow on a dial driven by gears. The cathode ray oscilloscope is an effective screen-based display for continuous analogue electrical signals.

Modern monitors can integrate various physiological measurements and display information in a variety of formats (e.g. discrete numbers, tables or waveforms). Most monitoring systems follow good ergonomic principles, with different variables separated consistently by position on the screen and by colour. Important information can be displayed in a variety of formats, much of which is modifiable by the user to display relevant information ( Fig. 17.3 ).

Fig. 17.3, Standard anaesthetic monitoring screen.

Despite many attempts to simplify patient data into geometric shapes or bar graphs, data continue to be displayed most often as simple numbers supported by waveforms (e.g. invasive pressures) and graphical or tabular (numeric) display of time trends. Trends are particularly useful when clinical problems may produce gradual change. For example, in neurosurgery a gradual decrease in end-tidal carbon dioxide concentration may indicate multiple air emboli.

Biological electrical signals

The detection and recording of biological electrical potentials are important clinical measurements that incorporate many of the key principles of clinical measurement.

Depolarisation of the cell membrane of excitable cells is fundamental to their action and generates a transient potential difference between the active cell and surrounding tissues. The summation of synchronous extracellular potentials from many excitable cells generates a widespread electric field detectable by electrodes on the body surface. The ECG and EEG are two well-established measures of biological electrical activity. Nerve conduction studies may be used to diagnose potential iatrogenic nerve injury.

Biological electrical signals are detected using electrodes constructed of silver and coated with silver chloride. Low, stable impedance (the resistance to alternating current) between skin and electrode minimises mains interference. Symmetrical electrode impedance and insignificant polarisation control drift. The electrolytic silver chloride layer is however very thin, prone to deterioration and only suitable for single use. Movement artefacts are minimised by separating the electrode surface from the skin with a foam pad impregnated with electrolyte gel. Degreasing with alcohol before applying the electrode helps reduce skin impedance and ensure satisfactory adhesion.

Biological electrical signals are recorded as waveforms. All complex waveforms can be described as a mixture of simple sine waves of varying amplitude, frequency and phase – Fourier analysis. These consist of a fundamental wave (the slowest sine wave in the waveform) and a series of harmonics that are multiples of the frequency of the fundamental wave and shifted in phase. The lower harmonics tend to have the greatest amplitude. For all continuous biological signals (including mechanical, e.g. arterial pressure waveform), a reasonable approximation may be obtained by accurate reproduction of the fundamental frequency and first 10 harmonics. It is important, however, that both the amplitude and phase difference of each harmonic are faithfully reproduced through the transduction system.

Amplification and gain stability

The amplitude of tiny bioelectrical signals must be increased by amplification and unwanted noise and interference minimised. The degree of amplification is termed gain and, whilst often user defined, should remain constant over the recording period. Calibration voltages may also be incorporated for correct adjustment of gain. The amplification of the signal should be constant across the whole range of signal amplitudes, and this amplitude linearity is often specified by equipment manufacturers for a specific amplitude range.

Input impedance and common mode rejection

Amplifiers for biological signals require high common mode rejection and high input impedance. The input and electrode impedances act as a potential divider: High electrode impedance and low amplifier input impedance are undesirable as they result in an attenuated electrical signal across the amplifier. The input impedance of modern amplifiers exceeds 5 MΩ to avoid problems, and careful attention must be paid to minimising electrode impedance, particularly for EEG recordings.

Pure amplification increases unwanted signals (noise) as much as the wanted – essentially it is the equivalent of the volume control on a radio. Differential amplification reduces unwanted noise. The potential difference between two input signals is amplified, but electrical signals common to both are attenuated. This feature is termed common mode rejection and effectively reduces mains interference in all biological signals and electrocardiographic contamination of smaller electroencephalographic signals. The common mode rejection ratio (CMRR) for a typical differential amplifier exceeds 10,000:1. In other words, a signal applied equally to both input terminals would need to be 10,000 times larger than a signal applied between them for the same change in output ( Fig. 17.4 ).

Fig. 17.4, Two types of amplifier; common mode rejection is particularly useful for minimising mains electricity interference. (A) Single-ended amplifier. (B) Differential amplifier.

Frequency response

The bandwidth of the amplifier describes the range of frequencies that it can accurately reproduce. It must cover the range of frequencies which are important in the signal. In practice, amplifiers require a flat frequency response for ECG from 0.14 to 50 Hz, for EEG from 0.5 to 100 Hz and for EMG from 20 Hz to at least 2 kHz.

Ideally, when there is no input signal, any measurement should be zero (zero stability). However, low-frequency interference, largely caused by slow fluctuating potentials generated in the electrodes, produces baseline instability and drift. This is removed through a network of resistors and capacitors functioning as a high-pass filter which allows higher frequency biological signals to pass but attenuates low-frequency noise. Whilst this reduces the bandwidth of the amplifier, such a compromise allows an interpretable signal free of baseline instability. For example, ECG monitoring requires a continuous recording with a stable baseline, which is achieved at the expense of waveform reproduction. High-pass filtering removes movement artefact, but attenuation of the low-frequency elements of the ECG, such as the T-wave, may cause distortion. Conversely, diagnostic ECGs require accurate reproduction of the waveform, but the long time constants of the amplifiers result in baseline instability, with movement artefact being a particular issue. Other filters can attenuate particular frequencies. Highly selective band-reject filters attenuate 50 Hz interference (main voltage) from the signal. Low-pass filters are used to eliminate higher-frequency artefacts from an EEG signal. The purpose of filtering is to reduce unwanted noise relative to the signal. When the frequency range of signal and noise overlap, some degree of signal degradation is inevitable ( Fig. 17.5 ).

Fig. 17.5, Both high- and low-pass filters can be positioned to minimise interference depending on the clinical requirements. The ECG components often overlap with noise from other sources, and attempting to filter these sources will necessarily affect the accuracy of the ECG recording.

Noise and interference

Electrical noise arising from the patient, the patient–electrode interface or the surroundings may interfere with accurate recording of biological potentials.

Noise originating from the patient

Millivolt ECG potentials on the body surface are hundreds of times larger than microvolt EEG signals on the scalp. Electromyographic signals may be even larger, and muscular activity, especially shivering, causes marked interference. Two features of electronic amplifier design substantially improve the EEG SNR. Electrocardiographic potentials are essentially the same across the scalp and are ignored by amplifiers with a high common mode rejection. Electromyographic activity has a higher frequency content than the EEG signal and may be minimised by a low-pass filter which attenuates the higher-frequency response of the amplifier to a level which attenuates the EMG signals and does not interfere with the characteristics of the EEG.

Noise originating from the patient–electrode interface

Recording electrodes do not behave as passive conductors. All skin–metal electrode systems employ a metal surface in contact with an electrolyte solution. Polarisation describes the interaction between metal and electrolyte which generates a small electrical gradient. Electrodes comprising a metal that is plated with one of its own salts (e.g. silver–silver chloride) avoid this problem because current in each direction does not significantly change the electrolyte composition. Mechanical movement of recording electrodes may also cause significant potential gradients – alteration in the physical dimensions of the electrode changes the cell potential and skin–electrode impedance. Differences in potential between two electrodes connected to a differential amplifier are amplified, and asymmetry of electrode impedance seriously impairs the CMRR of the recording amplifier.

Noise originating outside the patient

Electrical interference.

Mains frequency interference with the recording of biological potentials may be troublesome, particularly in electromagnetically noisy clinical environments. Patients function physically as large unscreened conductors and interact with nearby electrical sources through the processes of capacitive coupling and electromagnetic induction.

Capacitance permits alternating current to pass across an air gap. A live mains conductor and nearby patient behave as the two plates of a capacitor. The very small mains frequency current which flows through the patient is of no clinical significance but confounds the detection and amplification of biological potentials, creating unwanted interference in the recording. Capacitive coupled interference is minimised by reducing the capacitance and the alternating potential difference. This is achieved by moving the patient away from the source of interference and by screening mains-powered equipment with a conductive surround which is maintained at earth potential by a low-resistance earth connection and by surrounding leads with a braided copper screen – stray capacitances couple with the screen instead of the lead.

Alternating currents in a conductor generate a magnetic flux. This induces voltages in any nearby conductors which lie in the changing magnetic flux, including the patient or signal leads to the amplifier, which function as inefficient secondary transformers. This source of interference is minimised by keeping patients as far as possible from powerful sources of electromagnetic flux, especially mains transformers. Electromagnetic inductance may be minimised by ensuring that all patient leads are the same length, closely bound or twisted together until very close to the electrodes. This ensures that the induced signals are identical in all leads and therefore susceptible to common mode rejection.

The importance of low electrode impedance .

High electrode impedance may exaggerate the effects of surrounding electrical interference. Capacitive and inductive coupling produce very small currents in the recording leads. If the electrode impedance is low, the potential at the amplifier input must remain close to the potential at the skin surface so that minimal interference results. If electrode impedance is high the small induced currents may create a significant potential difference across that impedance, leading to severe 50-Hz interference.

Radio frequency interference from diathermy is a significant problem for the recording of biological potentials. Electrocardiographic amplifiers may be provided with some protection by filtering the signal before it enters the isolated input circuit, filtering the power supply to block mains-borne radio frequencies and enclosing the electronic components in a double screen, the outer earthed and the inner at amplifier potential.

Biological mechanical signals

Pressure measurements are employed widely in anaesthesia and critical care, using several physical principles and a range of instruments. Liquid column manometers display pressure relative to a predefined zero point using specific fluids of known density. Mechanical pressure gauges, used particularly in high-pressure gas supplies, rely on pressure-dependent mechanical movement being amplified by a gearing mechanism, which drives a pointer across a scale.

For most physiological pressure measurements, diaphragm gauges are used: a flexible diaphragm moves according to the applied pressure. Modern diaphragm gauges are sensed by a transducer that converts the diaphragm's mechanical energy into electrical energy. This is often displayed as a waveform and is subject to the same principles of waveform acquisition and amplification as described previously.

Electromechanical transducers

The first step is movement of the diaphragm proportional to applied pressure. This depends on the stiffness of the diaphragm and substantially determines the operating characteristics of the transducer. Linearity of amplitude and frequency response are improved by using small stiff diaphragms, but this requires a more sensitive mechanism for sensing diaphragm movement.

Wire or silicon-crystal strain gauges are based on the principle that stretching or compression of a wire or silicon changes the electrical resistance, capacitance or inductance. They are very sensitive and display an excellent frequency response, but non-linearity and temperature dependence are difficult technical problems.

Optical transduction senses movement of the diaphragm by reflecting light from the silvered back of the convex diaphragm onto a photocell. Applied pressure causes the silvered surface to become more convex. This causes the reflected light beam to diverge, reducing the intensity of reflected light sensed by the photoelectric cell. This design is used in fibreoptic cardiac catheters for intravascular pressure measurement. These miniature pressure transducers are expensive but have a high-frequency response, and fibreoptic light sources eliminate the risk of microshock.

The cardiovascular system

A core role for the anaesthetist is ensuring adequate oxygen supply to the patient's tissues. This can be estimated clinically: Capillary refill time, urine output and pulse volume are all clinical measurements that act as surrogates for tissue perfusion. Such clinical observations remain crucial despite the wide availability of electronic monitoring systems.

Electrocardiography

The electrocardiogram measures myocardial electrical activity and allows identification of heart rate, rhythm and abnormalities such as myocardial ischaemia or infarction. The synchronous depolarisation and prolonged action potentials in cardiac muscle summate to generate a potential field of relatively higher amplitude. This potential difference is detected between two electrodes placed on the body surface, and the third lead is used as a reference electrode. The very small absolute voltage changes (1 mV in amplitude with a frequency response of 0.05–100 Hz) require amplification before being displayed as a waveform.

Different lead positions detect electrical activity from different parts of the myocardium. The commonest position of the electrodes is the CM5 arrangement as this is the best position to detect ischaemia of the left ventricle ( Fig. 17.6 ).

Fig. 17.6, The CM5 configuration for electrocardiograph monitoring.

Alarms may be set to identify arrhythmias, bradycardia and tachycardia. Most monitors display the heart rate and, in addition, a measure of ST segment depression and elevation produced by cardiac ischaemia or infarction. This may be displayed as a trend over time. Unfortunately the relatively small voltages measured are easily swamped by skeletal muscle activity or surgical diathermy, leading to false alarms. The signal may also be severely degraded if the electrode gel has been allowed to dry out or if the weight of the leads pulls on the electrodes. The monitor only identifies ischaemia in a single area; multiple lead systems are required to monitor the whole myocardium.

Arterial pressure

Indirect methods

Aside from palpating the pulse, the majority of measurements of arterial pressure depend on signals generated by the occlusion of a major artery using a cuff, known as the Riva–Rocci method. Systolic pressure can be estimated by the return of a palpable distal pulse; auscultation of the Korotkoff sounds can determine systolic and diastolic pressures. These methods, however, are too time consuming during anaesthesia and difficult because of poor access to the patient's arm.

Oscillometric measurement of arterial pressure

Oscillometric measurement estimates arterial pressure by analysing the pressure oscillations produced by a cuff occluding pulsatile blood flow in an artery during cuff deflation. Modern machines use a single cuff with two tubes for inflation and measurement (commonly referred to as a DINAMAP – device for indirect non-invasive automatic mean arterial pressure). During slow deflation, each pulse generates a pressure change in the cuff distinguishable from the slowly decreasing cuff pressure. Above systolic pressure, these changes are small but suddenly increase in magnitude when the cuff pressure reaches the systolic point. As the cuff pressure decreases further, the amplitude reaches a peak and then starts to diminish. The point of maximal amplitude correlates closely with the mean arterial pressure. As the cuff pressure reaches diastolic pressure, the amplitude falls abruptly ( Fig. 17.7 ). To avoid high cuff pressures and long deflation times, monitors inflate the cuff to just above a normal systolic pressure and then slowly decrease the pressure until a pulse is detected. Consequently, estimates of diastolic pressure can be unreliable. If a pulse is not detected, the cuff is then inflated to a higher pressure. This process may be repeated several times before a measurement is made.

Fig. 17.7, The NIBP measurement uses a cuff that deflates in steps (line on graph); pressure oscillations from the vessels opening and closing increase in amplitude towards the mean arterial pressure, then cease when the cuff pressure is below diastolic.

Commercial instruments attempt to improve the reliability of the measurement. For example, at each successive plateau pressure during the controlled deflation, successive pressure fluctuations are compared and accepted only if they are similar. All automatic instruments require a regular cardiac cycle with no great differences between successive pulses. Accurate and consistent readings may be impossible in patients with an irregular rhythm, particularly atrial fibrillation. Furthermore, mechanical interference (e.g. patient shivering or movement of the arm) can prevent accurate measurement.

Clinical studies comparing automatic oscillometric instruments with direct arterial pressure have demonstrated good correlation for systolic and mean pressure with a tendency to overestimate at low pressures and underestimate at high pressures. Diastolic pressures are less reliable. The 95% confidence intervals approximate to 15 mmHg. The disadvantages of automated oscillometry are shown in Box 17.3 .

Box 17.3

Disadvantages of automated oscillometry

Delayed measurement with arrhythmias or patient movement

Inaccuracy with systolic pressure <60 mmHg

Inaccurate if the wrong size cuff used

May be inaccurate in obese patients

Discomfort in awake patients

Skin and nerve damage in prolonged use

Delay in injected drugs reaching the circulation

Backflow of blood into i.v. cannulae

Pulse oximeter malfunction as cuff is inflated

Other techniques

Alternative techniques use pressure, low-frequency sound, Doppler shift of an ultrasound signal or plethysmography. Most common amongst these in perioperative practice is the Penaz technique, which measures the effect of external pressure on the blood flow through a finger (e.g. Finometer, Nexfin, CNAP). The principle is of a control loop. The volume of the finger is measured using light absorption. A feedback loop to an encircling cuff provides counterpressure to the arterial pulsation to keep the volume constant. The pressure required to keep the volume constant is equivalent to arterial pressure. With improvements in technology, these devices are increasingly used for continuous non-invasive arterial pressure measurement. Some systems incorporate an automated recalibration process to counter issues of drift. Direct arterial cannulation is avoided, and accuracy is generally good.

Direct measurement

Direct arterial pressure measurement requires insertion of a cannula (20–22G) into an artery (usually radial because the ulnar artery may compensate for occluded radial flow) connected via a fluid column to the transducer. As fluids are incompressible, the pressure in the artery is transmitted to a transducer, which converts pressure into an electrical signal for display by the monitor. The transducer should be at the level of the left ventricle and the transducer opened to the atmosphere to provide a zero reading before use. Monitors usually display the waveform and systolic and diastolic pressures as well as the mean pressure, calculated by integration of the waveform to derive the average pressure across one cardiac cycle. The waveform provides useful additional information: visual estimates of pressure, frequency response and damping, and relative hypovolaemia during positive pressure seen as variability or swing in the waveform. Furthermore, it provides a visual approximation of contractility, vascular resistance and stroke volume ( Fig. 17.8 ).

Fig. 17.8, Arterial waveform for a single cardiac cycle; the slope of 1 approximates to cardiac contractility, 2 represents systolic pressure, 3 is the dichrotic notch (aortic valve closure), 4 is diastolic pressure and the area of 5 approximates stroke volume.

The main advantages of direct arterial pressure systems are that they provide accurate real-time measurement (essential when administering drugs such as vasopressors) and permit convenient blood sampling ( Box 17.4 ). Invasive arterial pressure monitoring has become standard practice for high-risk and severely ill patients, both in the operating theatre and ICU.

Box 17.4

Advantages of direct arterial pressure measurement

Accuracy of pressure measurement

Beat-by-beat observation of changes when blood pressure is variable or when vasoactive drugs are used

Accuracy at low pressures

Ability to obtain frequent blood samples

However, malpositioning of the transducer, failure to zero the transducer or problems with damping (see later) introduce error. For example, if the operating table is moved upwards while the transducer remains static, the difference in height artificially increases the pressure reading. Unusual readings should be checked against a reading from a non-invasive monitor. Problems relating to arterial cannulation are shown in Box 17.5 .

Box 17.5

Problems relating to arterial cannulation

Requires skill to insert

Bleeding

Pain on insertion

Arterial damage and thrombosis

Injury to nearby nerves (e.g. superficial radial nerve)

Embolisation of thrombus or air

Misplaced (retained) guidewire

Ischaemia to tissues distal to puncture site

Local infection and bacteraemia

Inadvertent injection of drugs

Late development of fistula or aneurysm

Resonant frequency and damping

As previously described, Fourier analysis converts complex waveforms into a series of sine waves. For an arterial waveform, the pulse frequency represents the fundamental frequency, with accurate reproduction of the waveform requiring transduction up to the 10th harmonic; that is, for a pulse of 120 bpm, transduction requires a linear frequency of up to (120 × 10) / 60 = 20 Hz. To ensure that the amplitude and phase difference of each harmonic is accurately reproduced, the transduction system must have a natural frequency greater than the frequency components of the waveform, as well as sufficient damping. The fluid and diaphragm of an arterial pressure transducer constitute a mechanical system which oscillates in simple harmonic motion at its natural resonant frequency. This determines the frequency response of the measurement system ( Fig. 17.9 ). If the resonant frequency of the measurement system overlaps any sine-wave component of the measured waveform, the entire system will have increased resonance (in this case, inaccurately elevated systolic pressure). The resonant frequency of a catheter-transducer measuring system is highest, and the damping effect of frictional resistance to fluid flow is lowest, when the velocity of movement of fluid in the catheter is minimised. This is achieved with a stiff, low-volume displacement diaphragm and a short, wide, rigid catheter.

Fig. 17.9, Stimulation of the output of a catheter-transducer system with increasing frequency of a constant amplitude input signal. The linearity of response is lost as the frequency approaches the resonant frequency of the system.

Determination of the resonant frequency and damping

The resonant frequency and the effects of damping may be estimated by applying a step change in pressure to the catheter-transducer system and recording the response ( Fig. 17.10 ). An underdamped system responds rapidly but overshoots and oscillates close to the natural resonant frequency of the system; frequency components of the pressure wave close to the resonant frequency are exaggerated. By contrast, an overdamped system responds slowly, and the recorded signal decreases slowly to reach the baseline, with no overshoot. High-frequency oscillations are damped, underestimating the true pressure changes. These extremes are undesirable. In general, MAP is least affected by damping, but an underdamped system will overestimate systolic pressure and underestimate diastolic pressure. The converse is true for an overdamped system: The systolic pressure will be an underestimate, and the diastolic will be an overestimate.

Optimal damping

Optimal damping maximises the frequency response of the system, minimises resonance and represents the best compromise between speed of response and accuracy of transduction. A small overshoot represents approximately 7% of the step change in pressure, with the pressure then following the arterial waveform (see Fig. 17.10 ).

Damping is relatively unimportant when the frequencies being recorded are less than two thirds of the natural frequency of the catheter-transducer system. Modern transducer systems using small compliance transducers connected to a short, stiff catheter with a minimum of constrictions or connections approximate to this ideal. The system also includes a pressurised bag of 0.9% saline producing a flow of 1–3 ml h ^–1 through a restrictor to prevent clot formation, as well as allowing a higher flow rate to flush the system, such as after blood sampling. Air bubbles in the system, clotting or kinking in the vascular catheter and arterial spasm lower the natural resonant frequency and increase the damping.

Accuracy of arterial pressure measurements

Invasive devices provide the accepted gold standard for arterial pressure measurement. However, the catheter-transducer system requires careful setup, and arterial pressure varies throughout the arterial tree. As the pulse wave travels from the ventricle to peripheral arteries, changes in vessel diameter and elasticity affect the pressure waveform, which becomes shorter with increased amplitude. Differences in arterial pressure between limbs are common, particularly in patients with arterial disease.

Indirect methods using an occluding cuff make intermittent measurements, with the systolic and diastolic readings reflecting the conditions in the artery at the time they are measured. By contrast, direct pressure measurements are the average of several cycles, more precisely reflecting mean pressure. Indirect measurements may be compromised by taking a small number of infrequent samples from a variable signal.

Central venous pressure

Central venous pressure represents the pressure of blood entering the right atrium, usually 2–3 mmHg. Whilst often considered a measure of central venous blood volume, it is notoriously unreliable in the critically ill or those with abnormal right-sided cardiac function, and alone it rarely provides an accurate reflection of a patient's circulating fluid volume. However, access to the central venous system is useful for the administration of vasoactive or irritant drugs or for central venous sampling.

There are four common routes for central venous catheterisation.

Internal jugular catheters are used most commonly because the vein is superficial, of larger diameter and easily managed. This is often the most appropriate route for use in an emergency. However, the insertion point is adjacent to several vital structures, including the carotid artery, pleura, brachial plexus and cervical spine, risking direct needle trauma to these structures. Ultrasound guidance is recommended for internal jugular catheterisation.
Catheters inserted into the subclavian vein offer similar problems to internal jugular venous catheters, and if accidental arterial puncture occurs, the overlying clavicle obscures bleeding and makes direct compression of the artery impossible. The proximity of the pleura increases the risk of accidental lung puncture. The subclavian route should therefore be used only after first considering the internal jugular approach. The insertion point under the clavicle may make it easier to anchor the catheter to the skin, which is an advantage for longer term use. Ultrasound guidance can assist insertion but is technically more challenging than for internal jugular access.
Long, small-diameter catheters inserted via the antecubital fossa are relatively easy and safe to insert. However, catheters inserted via the basilic or cephalic vein are sometimes difficult to advance past the shoulder. X-ray imaging is required to confirm placement, and thrombosis of the veins is common after 24 h. The length of the narrow catheter reduces frequency responsiveness.
Femoral venous catheters are inserted just below the inguinal ligament. They are relatively easy to insert and may be of large gauge to allow rapid transfusion of fluids. This route is often chosen in children. However, the site of insertion is often within a skin fold, making skin flora contamination more likely.

Central venous catheters are usually connected to the same type of transducer and flush system described for arterial cannulae. This provides a continuous pressure reading. However, because central venous pressure (CVP) is low, great care is required to ensure that the pressure is measured relative to the correct zero point (fourth intercostal space in the midaxillary line) on the patient ( Fig. 17.11 ). Single readings may help diagnose right-sided cardiac failure such as after acute pulmonary embolus or cardiac tamponade. Repeat measurements were traditionally used to guide fluid therapy, but this should be discouraged as there is little correlation between CVP and cardiac output in response to fluid challenges. Complications ( Box 17.6 ) are infrequent but potentially serious.

Fig. 17.11, Surface markings used to identify the position of the right atrium.

Box 17.6

Complications of central venous catheterisation

Acute

Arrhythmias

Bleeding

Air embolus – on insertion or removal

Pneumothorax

Damage to thoracic duct, oesophagus, carotid artery, stellate ganglion

Cardiac puncture

Catheter embolisation

Misplaced (retained) guidewire

Delayed

Bacteraemia/sepsis

Thrombosis

Cardiac rupture

Pulmonary artery pressure

Cardiac output and systemic arterial pressure are determined by the filling pressure of the left side of the heart. The pulmonary artery flotation catheter (PAFC, or Swan-Ganz catheter) ( Fig. 17.12 ) enables measurement of left-sided pressures, imperfect surrogates for left ventricular end-diastolic volume. It is rarely used now because of frequent complications, a lack of evidence of improved survival and the introduction of non-invasive techniques to estimate cardiac output.

Fig. 17.12, Distal end of a pulmonary artery catheter showing inflated balloon and thermistor.

A PAFC is a long catheter with three or four lumens and a thermistor near the tip. It is inserted into the internal or subclavian vein through a large cannula. A flexible plastic sheath allows the catheter to be inserted, withdrawn and rotated after insertion without desterilising it. After insertion into the superior vena cava, saline is injected to inflate a balloon at the tip. The pressure at the tip is measured via a transducer and displayed on a monitor. The catheter is then advanced slowly so that the blood flow directs the catheter towards the pulmonary artery. As the catheter is advanced, a series of changes in pressure is observed, marking the progression through the right atrium and right ventricle into the pulmonary artery ( Fig. 17.13 ). Eventually the balloon wedges into a pulmonary artery. At this point, the tip is isolated from the pulmonary artery and measures the pressure in the pulmonary capillaries, a reflection of left atrial pressure. Although the ability to estimate left atrial pressure is useful, repeat measurements after a circulatory challenge are more informative than a single reading.

Fig. 17.13, Diagrammatic representation of pressure waveforms. R, Right.

The PAFC has led to many advances in our understanding of cardiac physiology and the mechanisms and treatments of diseases such as sepsis. However, catheter placement may be difficult, and prolonged manipulation may be needed to direct the catheter into the pulmonary artery. Arrhythmias are extremely common during catheter insertion, and the technique carries specific risks ( Box 17.7 ) in addition to all those of central venous catheterisation (see Box 17.6 ).

Box 17.7

Risks of pulmonary artery catheterisation (in addition to those in Box 17.6 )

Arrhythmias with catheter manipulation

Damage to tricuspid and pulmonary valves

Knotting of catheter

Pulmonary infarction if balloon left inflated

Pulmonary artery rupture with balloon inflation

Cardiac rupture

Cardiac output

Cardiac output is closely linked to oxygen delivery; in clinical practice, low cardiac output is linked to increased mortality. However, factors such as the pressure changes caused by positive pressure ventilation, changes in heart rate and arrhythmias introduce complexities in measurement and interpretation. Therefore monitors do not produce consistent results even when synchronised with the heartbeat and respiratory cycle. Dilution techniques have been regarded as the gold standard against which other methods are compared.

The Fick principle

The Fick principle defines flow by the ratio of the uptake or clearance of a tracer within an organ to the arteriovenous difference in tracer concentration. It may be used to measure cardiac output (oxygen uptake or indicator dilution) and regional blood flow (e.g. cerebral blood flow using the uptake of nitrous oxide) and renal blood flow from the excretion of compounds cleared totally by the kidney, such as para-aminohippuric acid.

In patients with minimal cardiac shunt and reasonable pulmonary function, pulmonary blood flow may be calculated from the ratio of the oxygen consumption and the difference in oxygen content between arterial and mixed venous blood, as follows:

Pulmonary blood flow = \frac{oxygen consumption}{arteriovenous oxygen content difference}

$Pulmonary blood flow = \frac{oxygen consumption}{arteriovenous oxygen content difference}$

Oxygen consumption from a reservoir is measured using an accurate spirometer and oxygen content measured with a co-oximeter. Measurements should be made at steady-state situations, with constant inspired oxygen concentration and blood samples obtained slowly whilst the oxygen consumption is being determined. True mixed-venous blood samples must be obtained from a pulmonary artery catheter. Alternative indicator dilution techniques described here are less demanding. The effects of ventilation and beat-to-beat variation in cardiac output are averaged over the long period of measurement of oxygen consumption. Errors in measurement of oxygen consumption limit the accuracy of this technique (±10%).

Indicator dilution

An indicator is injected as a bolus into the right side of the heart, and the concentration reaching the systemic side of the circulation is plotted against time ( Fig. 17.14 ). The average concentration is calculated from the area under the concentration–time curve divided by the duration of the curve. The cardiac output during the period of this measurement is the ratio of the dose of indicator to the average concentration.

The general formula is:

\begin{matrix} Cardiac output (litres m i n^{- 1}) \\ = \frac{indicator dose (units) \times 60}{average concentration (units litr e^{- 1}) \times time (s)} \end{matrix}

$\begin{matrix} Cardiac output (litres m i n^{- 1}) \\ = \frac{indicator dose (units) \times 60}{average concentration (units litr e^{- 1}) \times time (s)} \end{matrix}$

The main problem with this technique is that when the dye has been measured at the artery, it passes back to the heart and then back to the arteries (known as recirculation), making the calculations more complex. This may be circumvented by extrapolation of the early exponential downslope to define the tail of the curve which would have been recorded if recirculation had not occurred (see Fig. 17.14 ). The area under the curve is calculated by integration.

Chemical indicator dilution

Original studies used indocyanine green, a non-toxic chemical indicator with a relatively short half-life. It also has a peak spectral absorption at 800 nm, the wavelength at which absorption of oxygenated haemoglobin is identical to that of reduced haemoglobin. The measurement is therefore not affected by arterial saturation. However, because the dye is cleared from the circulation only slowly, recirculation makes repeated measurements impossible. More recent monitors use an injection of lithium as a marker, which is detected with a modified arterial catheter.

Thermal indicator dilution

This technique requires a PAFC to be in the pulmonary artery. The principle is similar to other indicator dilution methods. A bolus of 10 ml cold saline is injected into the right atrium, and a thermistor at the tip of the PAFC measures the temperature change. The smaller the temperature drop, the larger the cardiac output. The recorded temperatures generate an exponential dilution curve with no recirculation. The heat dose is the difference in temperature between the injectate and blood multiplied by the density, specific heat and volume of the injectate. The average change in heat content is the area under the temperature–time graph multiplied by the density and specific heat of blood. However, errors occur where mixing of the fluid bolus with venous blood is incomplete, and correction values are necessary, such as for changes in injectate temperature during injection through the catheter. The average of three separate readings is taken as a reliable cardiac output measurement and should be repeated, particularly after changes in therapy.

The measurement of cardiac output using both dye and thermodilution is now automated with computer-controlled sampling, calculation of indicator dilution curves, rejection algorithms for artefacts or curves which are not exponential and online calculation of cardiac output.

Continuous cardiac output monitors have also been introduced. These use a similar principle, but instead of using a bolus of cold saline, the catheter has an electrical coil which is heated at intervals, creating a bolus of warm blood that passes into the pulmonary artery. This eliminates much of the operator error and produces frequently updated measurements of cardiac output, allowing the effect of interventions to be observed.

Monitors have also been developed that do not require a pulmonary artery catheter but use a bolus of iced saline injected into a modified central venous catheter and a peripheral arterial catheter with a built-in thermistor (e.g. the PiCCO (pulse index continuous cardiac output) monitor).

Most of the current monitors allow cardiac output data to be integrated with other measurements such as arterial and venous pressure to provide calculated values of, for example, systemic vascular resistance and stroke volume. This aids the choice and administration of vasoactive drugs.

Pulse contour analysis

The shape of the arterial pulse (the pulse contour) is a product of the rate of ejection of blood into the aorta and the elasticity of the arterial tree. Therefore, if some assumptions are made about the arterial tree, the volume ejected at each heartbeat (stroke volume) may be calculated from the shape of the arterial pulse contour. Multiplying this by the heart rate provides an estimate of cardiac output. This approach has the advantage of being able to calculate the cardiac output in near real-time using an arterial cannula alone.

However, the technique relies on assumptions on arterial tree elasticity which may not always be true in every patient. Therefore these systems may require calibration by another method such as thermodilution every 8–12 h to ensure accuracy (e.g. PiCCO and LiDCO (lithium dilution cardiac output)) use thermodilution and lithium, respectively, for calibration). Recently developed devices use complex algorithms linked with sensor measurements to make estimates of arterial elasticity and vascular tone and are promoted as not requiring calibration (e.g. FloTrac). Both the calibrated and uncalibrated systems also give a value for the variation of stroke volume with respiration, comparable to the arterial pressure swing described earlier. Higher values predict fluid responsiveness, provided certain conditions are met (e.g. closed chest, mandatory ventilation, tidal volumes of 8 ml kg ^–1 ). However, the effects of critical illness and vasopressor use influence the accuracy and precision of these pulse contour analysis systems, and in general they are inferior to PAFC thermodilution techniques. These systems are commonly used as part of goal-directed fluid therapy (see Chapter 30 ).

Doppler ultrasonography

Ultrasound techniques can detect the shape, size and movement of tissue interfaces, especially soft tissues and blood, including the echocardiographic measurement of blood flow and the structure and function of the heart. Sound waves are transmitted by the oscillation of particles in the direction of wave transmission, defined by amplitude (the difference between ambient and peak pressures) and the wavelength (distance between successive peaks) or frequency (the number of cycles per second) (see Chapter 15 ). These characteristics are measured by a pressure transducer placed in the path of an oncoming wave. The human ear detects frequencies within the range of 20–20,000 Hz. Diagnostic ultrasound uses frequencies in the range of 1–10 MHz. Short-term diagnostic use of ultrasound appears to be free from hazard.

Generation and detection of ultrasound

The physics of ultrasound and Doppler effects are explained in Chapter 15 .

Properties of ultrasound

Shorter wavelengths and higher frequencies improve resolution but reduce tissue penetration. Amplitude determines the intensity of the ultrasound beam, the number and size of echoes recorded and therefore sensitivity. Ultrasound is absorbed by tissues and reflected back at the junction between two tissues, tissue–fluid or tissue–air. This reflection at interfaces is the basis for diagnostic use of ultrasound. The intensity of the beam decreases exponentially as it passes through tissue. Attenuation depends on the nature and temperature of the tissue and is related linearly to the frequency of the ultrasound.

Reflections at most soft-tissue interfaces are weak, but bone–fat and tissue–air interfaces reflect the majority of incident energy. Structures lying behind a bone or air interface cannot be studied using ultrasound. Various ultrasound techniques are suited to different applications and have extremely sophisticated two-dimensional, real-time, brightness- and colour-modulated displays under microprocessor control.

Detection of motion by the Doppler effect: cardiac output

When ultrasound waves reflect off an object moving towards the transmitter, there is an apparent increase in frequency as the object encounters more oscillations per unit time. This is termed the Doppler effect. The change in frequency is proportional to the velocity of the object and two constants: the frequency of the transmitted ultrasound and the velocity of ultrasound in the medium. The velocity ( v ) of the object can be calculated using the Doppler equation:

v = \frac{c f_{d}}{2 f_{t} \cos θ}

$v = \frac{c f_{d}}{2 f_{t} \cos θ}$

where f _d is change in Doppler frequency, c is speed of sound in medium, f _t is transmitted frequency, and θ is angle of probe relative to the flow of blood.

In practice a beam of ultrasonic waves is focused on the descending aorta, and reflections from red cells are measured by a transducer in the same probe. Probes may be transthoracic (usually placed in the sternal notch) or placed in the oesophagus. The signal obtained is displayed on the screen and indicates peak velocity and flow time. By making a number of assumptions about the nature of flow in the aorta, the cross-sectional area of the aorta (estimated from body surface area and age with a nomogram) and the percentage of CO passing down the thoracic aorta, SV and CO can be estimated. The internal algorithm for these calculations is based on calibration of total left ventricular stroke volume (measured by a PAFC) against descending aortic blood flow velocity and stroke distance as measured by the ODM ( Fig. 17.15 ).

Fig. 17.15, Display of an oesophageal Doppler cardiac output monitor showing the pulse waveform from which a variety of haemodynamic variables are derived.

The advantages of these monitors are that they produce an almost real-time estimate of cardiac output, stroke volume and other calculated cardiovascular variables. Disadvantages include that they rely on the ultrasound beam being directed at the centre of the aorta and assume the aorta is a smooth tube. In practice, even small movements of the sensor may lead to marked changes in readings because the speed of red cells near the aortic wall is measured. Furthermore, the aorta is not completely circular and may contain atheroma, and the diameter (an estimate) can change by as much as 12% during systole. The Doppler shift also depends on the direction of the ultrasound beam relative to the axis. Provided that the angle is less than 20 degrees, the error in cardiac output is only about 6%. Conscious patients do not always tolerate the more accurate oesophageal probes, and certain surgical procedures preclude their use (e.g. oesophagectomy). Most systems provide a visual (and audible) measure of signal strength to assist in focussing the probe.

Despite these limitations, the ODM may be useful in high-risk patients undergoing relatively minor surgery or in patients undergoing surgery requiring cardiovascular control (see Chapter 30 ). However, whilst useful for monitoring responses to treatment, they are unable to provide reliable estimates of actual cardiac output. Complications are few.

Transoesophageal echocardiography

Transoesophageal echocardiography (TOE) uses a miniaturised ultrasound probe inserted into the oesophagus under anaesthesia. It provides a real-time picture of all four cardiac chambers and valves. Advantages are that it can identify any malfunctioning valves, any ischaemia-induced wall motion abnormalities and whether therapy has been successful (e.g. valve surgery). However, the equipment is expensive and requires an operator trained in use and interpretation of the data. In cardiac surgery they can demonstrate adequate cardiac function and valve competence at weaning from cardiopulmonary off bypass. They can also measure cardiac output using Doppler. However, they are only suitable for anaesthetised or sedated patients, cannot provide prolonged continuous measurements and are rarely used in non-cardiac surgery. Transthoracic echocardiography is increasingly employed in ICU to answer specific haemodynamic questions, such as cardiac contractility, identification of regional wall motion abnormalities and ejection fraction. Additional information about the lungs and pleura can be obtained at the same sitting. However, accurate results are operator dependent, and continuous cardiac output monitoring is not possible.

Thoracic electrical bioimpedance

Tissue impedance depends on blood volume. Measurement of thoracic impedance provides an index of stroke volume. Two circumferential electrodes are placed around the neck and two around the upper abdomen. A small (<1 mA), constant, high-frequency (>1 kHz) alternating current is passed between the outer electrodes, and the resulting potential difference is detected by the inner pair. This potential is rectified, smoothed and filtered to record voltage fluctuations which reflect changes in impedance as a result of ventilation and cardiac activity. The cardiac activity is extracted by signal averaging relative to the ECG R-wave. This represents changes in thoracic blood volume and clearly resembles the pulse waveform.

Modern instruments show a modest agreement with invasive measurements of cardiac output, although trends and rapid changes in cardiac output are reliably demonstrated. This method is inaccurate when there are intracardiac shunts or arrhythmias and underestimates cardiac output in a vasodilated circulation.

The respiratory system

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here