Noninvasive temporary pacemakers and defibrillators are useful devices for treating arrhythmias that may arise during surgery and anesthesia as well as in intensive care or resuscitation settings. Although these are not specifically anesthesia devices, it is important for the anesthesiologist to understand the principles of their function and use.
∗ The author was a principal developer of noninvasive temporary pacemakers at Zoll Medical (Redmond, WA).
Clinical pacing, first introduced by Paul M. Zoll in 1952, used a device very similar to modern noninvasive pacers. However, it was not well accepted because it was a radical innovation in therapy in that it anticipated subsequent developments in monitoring, medical electronics, and intensive care. Although safe and effective, the technique had problems, such as interference with the electrocardiogram (ECG) signal, which made it difficult to determine effectiveness. In addition, the pacing stimulus produced severe pain. Although useful for resuscitation, noninvasive pacing was not satisfactory for long-term pacing therapy.
Implantable pacemakers followed the development of transistors and tissue-compatible stimulating electrodes. Problems of wire breakage, battery life, and circuit reliability were resolved by about 1980. Following the introduction of transvenous pacing, temporary transvenous pacing became popular for resuscitation, prophylaxis, and short-term pacing. However, in certain situations, the transvenous approach has serious disadvantages. Most important, it is not available rapidly enough for resuscitation of an unexpected cardiac arrest, and in such a situation, it is cumbersome and highly unreliable. As an invasive technique, transvenous pacing carries risks such as hemorrhage, infection, tamponade, embolization, and arrhythmias. For prophylaxis, the need for pacing must be weighed against both the expense of the procedure and the risk of complications.
As with all excitable tissue, electrical stimulation of cardiac muscle requires forcing decreases in transmembrane potential to activate voltage-dependent ion channels. Subthreshold stimuli have only transient and local effects. Suprathreshold stimuli result in action potentials that propagate throughout the cell membrane and, in the case of cardiac muscle, throughout the contiguous muscle mass. If atrioventricular (AV) conduction is intact, propagation is throughout the heart. Stimuli may be introduced either with transmembrane microelectrodes or by current fields from macroscopic electrodes, such as transvenous or epicardial pacing electrodes or external electrodes on the chest wall. Pacing electrodes in direct contact with tissue, on surface areas of a fraction of a square millimeter, require only 0.1 to 1.0 mA of current for effectiveness. With long-term electrodes, the development of 1-mm thickness or more of scar tissue increases their effective size and may require several milliamperes of current for stimulation. Electrodes on the chest wall are several centimeters distant from the myocardium and typically require 100 times as much threshold current. With an external device, unlimited power is available to ensure effectiveness. However, large currents passing through the skin, subcutaneous tissue, and muscle may cause considerable discomfort and even burns. The hemodynamic effects are similar to epicardial or endocardial ventricular pacing.
Improvements in the comfort of noninvasive temporary pacing have contributed greatly to the renewed interest in this technique. The original device of 1952 usually caused unacceptable pain in conscious patients, even following sedation and analgesia. Compared with a typical pacing threshold of 150 mA, a sharp stinging pain was produced at 17 mA. With newer devices, the pacing threshold has been reduced relative to the threshold for discomfort by modifications to the electrodes and stimulus waveforms. Pacing is usually well tolerated for at least a few minutes at currents of 80 mA or more, even without sedation or analgesia. Pacing thresholds vary from 31 mA to more than 100 mA and average about 55 mA. Sedation or analgesia usually improves acceptance by the patient, especially when pacing is required for several minutes.
The modern noninvasive temporary pacemaker system consists of a pacemaker pulse generator incorporated into a portable ECG monitor (most models include a portable defibrillator as well) and disposable self-adherent electrodes for the anterior and posterior chest wall ( Fig. 19-1 , A and B ). A three-lead ECG monitors effective ventricular capture and is required for the demand ventricular (VVI) pacing mode. The pacemaker has three main controls: the on/off switch, the rate selector, and the current amplitude selector. For operation, the electrodes are applied as in Figure 19-2 . A pacing rate (or escape rate in VVI mode) is selected according to clinical criteria, and the pacing current’s amplitude is gradually increased until ventricular capture occurs.
Theoretically, there should be an optimum size of electrode. At one extreme, the current field of a point source falls off as the inverse square of the distance from that point. Accordingly, a very small electrode on the chest wall would behave like a point source at a distance from the heart. At the other extreme, a very large electrode will have a uniform current field and will stimulate every muscle in the body, including the heart. A range of sizes and positions for electrodes was tested in a variety of patients to find the optimal size that would be accepted by most patients and for which precise placement was unnecessary (see Fig. 19-2 ). For the anterior electrode, a diameter of 10 cm was found to be ideal. The posterior electrode is larger and serves as a return path for the current. More recently, computational models have supported the size, position, and impedance determined experimentally. The negative polarity of the stimulus is applied to the anterior electrode for the lowest threshold.
An additional step was taken to explore the premise that most of the discomfort from noninvasive pacing was related to current density in the skin and subcutaneous tissues. This component of pain, felt subjectively as a sharp or stinging sensation, is different from pain associated with strong muscle contraction, which also may be uncomfortable. Figure 19-3 shows that the threshold current for skin pain, expressed as current density at the skin, is independent of electrode area, except for the smallest electrode, where edge effects or coarseness of sensory organs are important. The cardiac stimulation threshold is shown for comparison.
To reduce pain, current density at the skin should be reduced. This can be accomplished by ensuring that the current distribution across the electrode is uniform. Electrodes of relatively high impedance that match skin impedance help reduce edge effects and avoid hot spots if the skin itself is nonuniform. If practical, perspiration or ECG paste should be cleaned from the skin to achieve the most comfortable stimulation. The skin should not be shaved before the electrodes are applied. Because of the high impedance, up to half the energy supplied by the noninvasive external pacemaker may be dissipated in the electrodes. The Zoll NTP (Zoll Medical Corporation, Chelmsford, MA) is designed to supply selected current independent of impedance up to 2000 Ω. However, if the same electrode is to be used for monitoring or defibrillation, a low-impedance medium is necessary.
Electrodes are placed to avoid large muscles yet maintain proximity to the myocardium. The optimum location for the anterior electrode can vary considerably among individuals, and some experimentation may be necessary to achieve greater comfort. Otherwise, the electrode should be placed just to the left of the sternum and mostly below the pectoral muscle. Stimulation directly over the sternum usually is uncomfortable. Precise positioning of the posterior electrode is not necessary.
The noninvasive pacemaker uses the fact that cardiac muscle responds to stimulation somewhat differently than either nerves or skeletal muscle, which respond to stimuli of much shorter duration. Stimuli that are long and uniform in time can reliably pace with less discomfort. Figure 19-4 shows normalized threshold curves for muscle stimulation and for pacing as a function of stimulus duration. Increasing the stimulus duration from 0.5 or 2.0 ms to 40 ms greatly reduces the cardiac stimulation threshold. At 40 ms, threshold amplitude is 5% less than at 20 ms, and discomfort is subjectively halved. Although stimuli of long duration feel different, they are no more intense. Again, the controlled current source design of the Zoll NTP produces a long, constant-amplitude current pulse despite the complex impedance of the tissue and electrodes ( Fig. 19-5 ). The new stimulus does have the disadvantage of increased interference with ECG monitors, and questions of safety have been raised. Noninvasive external pacemakers with shorter stimulus duration and less uniform current amplitude may be more uncomfortable in conscious patients.
The most serious potential complication of pacing is that of precipitating arrhythmias. Early electrophysiologic studies implicated stimuli of long duration in the production of arrhythmias. However, the original noninvasive pacer used a waveform of 2 to 3 ms and probably never precipitated arrhythmias. By comparison, various permanent and temporary pacers have produced constant voltage or constant current waveforms from 0.2 to 4.5 ms in duration. Demand modes often are used in pacemakers, in part for their hemodynamic benefits, but also because of concern about stimulating during the relative refractory, or “vulnerable,” period, when the myocardium is partially repolarized, and a large, suprathreshold stimulus might precipitate ventricular tachycardia or fibrillation. Indeed, when permanent pacing electrodes with small surface areas were introduced to reduce threshold and extend battery life, sudden death occasionally resulted when they were used with older pulse generators that produced higher outputs. Especially when the patient has ischemia or myocardial damage from a recently placed electrode and consequent electrophysiologic inhomogeneity, very large stimuli during the relative refractory period should be avoided. On the other hand, operation of the noninvasive temporary pacer near threshold in the relative refractory period does not produce arrhythmias and is merely ineffective. The maximum output, only four times threshold in the most sensitive patients, is not likely to cause arrhythmias unless the patient’s myocardium is so unstable that any extrasystole could degenerate to an arrhythmia. Because the purpose of the device is to produce extrasystole, this risk is unavoidable. Mechanical stimulation from temporary pacing wires occasionally triggers arrhythmias, and this cause is avoided as well. Certain patients who have fibrillated when a temporary wire was passed in the setting of acute myocardial infarction have done well with noninvasive temporary pacing.
The effect of the long-duration stimulus on safety has been studied in dogs. A safety factor (i.e., a therapeutic ratio) was defined as the fibrillation threshold divided by the pacing threshold, measured at various stimulus durations. To obtain the worst-case result, the stimulus was intentionally applied during the relative refractory period for fibrillation measurements. Episodes of fibrillation occurred at 5 to 10 times the pacing threshold or higher, and no change was apparent in the safety factor at 40 ms compared with 2 ms.
One significant disadvantage of the long stimulus duration is its interference with ECG monitors. The stimulus artifact is much larger than ECG voltages (1000 times or more) and is within the normal ECG bandwidth. This artifact overloads most monitors, obscuring any real signal. Some monitors include overload recovery circuits and will function normally in this situation. However, the noninvasive temporary pacer is supplied with a rhythm monitor that has preamplifier circuits designed to accommodate the pacing stimulus. The pacemaker circuit inserts a marker on the ECG display to indicate at which point in the cardiac cycle the stimulus occurs. This marker consists of a 40-ms square wave followed by a 40-ms isoelectric period. Following these, the real ECG signal resumes. Part of the QRS response typically is visible as well as the T wave ( Fig. 19-6 ). On occasion, it is necessary to switch leads to obtain a clear tracing. The ECG monitor and pacing output are electrically isolated in accordance with the Association for the Advancement of Medical Instrumentation (AAMI) standards.
The role of noninvasive external pacing in cardiac arrest situations is under investigation. Fibrillation as a mechanism of arrest is probably more common than asystole, although asystole may follow successful defibrillation. Asystole carries a worse prognosis, in part because it often occurs at a late stage in prolonged arrest. However, some patients are certainly still viable and might be saved with pacing. Some early studies examining the effectiveness of pacing in arrest situations found that pacing held no advantage except in bradycardia. These studies were probably not sufficiently sensitive to discriminate between the situations in which pacing could be effective and those in which the patient was beyond salvage. More recent studies have found pacing to be of benefit in prehospital arrest situations.
Similar to transvenous pacing, noninvasive pacing can be used to interrupt and terminate most supraventricular and ventricular tachycardias using single stimuli, bursts, or overdrive pacing. For tachycardias over 180 beats/min, a special rate generator is needed and is easily added. Often all that is required to interrupt the circus movement in a segment of its AV junctional or ventricular path is a single ectopic ventricular beat.
Another potential application of noninvasive pacing is stress testing. As with transvenous pacing, noninvasive temporary pacing at rapid rates simulates exercise stress. This technique does not involve the risks accompanying the insertion of a temporary wire. Stress is alleviated immediately upon cessation of pacing and thus enhances the safety of the procedure. The pacing is interrupted briefly to look for ischemic changes on the ECG. Although rapid noninvasive pacing may be less comfortable than pacing at normal rates, it usually can be performed with analgesia. The effectiveness of the Zoll NTP for stress testing has been confirmed, but extensive experience, and especially correlation to standard protocols, is lacking.
For pediatric use, small electrodes are used for infants and children weighing less than 15 kg. Although pacing thresholds are the same for both children and adults, pediatric pacing rates are likely to be two to three times higher than those for adults. For sick or premature neonates, prolonged use may cause burns if the skin is immature and poorly perfused. However, brief use while securing other means of pacing is safe.
In the Operating Room
Currently, most patients who have significant conduction or sinus node dysfunction have permanent pacemakers. Although bradycardia or asystole unresponsive to atropine or other chronotropic agents is unusual, it has been reported. Temporary pacing frequently is used during the placement of a pulmonary arterial catheter when the patient has preexisting left bundle branch block. Although complete heart block during catheter placement probably occurs in fewer than 1% of cases, it can be disastrous, and noninvasive pacing is ideal for this situation. A temporary transvenous pacemaker requires another site of central access and is not without risk.
Noninvasive pacing is useful for any procedure involving permanent pacemaker systems. It can ensure uninterrupted rhythm during initial implantation of a pacemaker or during revisions or changes of a pulse generator, when mechanical problems occasionally lead to interruption of pacing. The electrodes for the NTP are nearly transparent to radiographs and do not interfere with fluoroscopy.
For patients with permanent pacemakers, the noninvasive pacing guarantees pacing capability during surgery if the system is damaged or reprogrammed by electrocautery. Although uncommon, intraoperative failure of a pacing system occasionally is reported. During surgery, electrocautery usually interferes with the sensing function of the noninvasive pacing, just as it does with permanent pacing systems.
In the event of unexpected arrest, noninvasive pacing can be applied within seconds. When applied promptly, it usually is effective in producing an electrophysiologic response, even if contractility is inadequate. The presence of an electrophysiologic response to pacing stimuli is the appropriate measure of effectiveness, and other pacing modalities produce no additional hemodynamic benefit.
Defibrillation, introduced into clinical practice by Beck and colleagues in 1947, used electrodes applied directly to the heart. Thoracotomy was advocated as the standard protocol for cardiac arrest. In 1956, Zoll and colleagues introduced closed-chest defibrillation with an alternating current (AC) waveform. Lown and Edmark introduced the direct current (DC) defibrillator in the early 1960s. In the early 1970s, Mirowski and colleagues introduced automatic implantable cardioverter-defibrillators (ICDs) for the treatment of patients with otherwise intractable ventricular arrhythmias. Since that time, there has been renewed interest in other waveforms that might reduce the energy required for defibrillation. Waveforms with improved effectiveness and safety have been discovered and put into practice. Biphasic waveforms with a fixed, opposite polarity component have largely replaced the underdamped oscillator technology.
The basic principle of defibrillation is to interrupt random and chaotic electrical activity in the heart by using a large stimulus that excites a large and sufficient fraction of the muscle mass at once. However, the detailed mechanism remains extremely controversial. A number of theories and observations have been proposed. One study in dogs found that one small area of fibrillating tissue (approximately 25%) may remain (after treatment)without the fibrillation becoming generalized again; however, two such areas usually lead to clinical failure to defibrillate. It has been observed that the defibrillation threshold may be related to the upper limit of vulnerability for producing fibrillation, although the significance of this is unclear. Much of the myocardium is in the refractory state during fibrillation; therefore a prolongation of recovery may be involved in the defibrillation mechanism. Unlike pacing, which requires an excitement stimulus of only a single cell, a defibrillation shock must excite a considerable portion of the myocardium, even though part of it may be in a relative refractory state, and 1000 times as much current may be required. Cardioversion of atrial or ventricular tachycardia may require much less current, and a pacing stimulus that produces a single extrasystole will occasionally suffice.
On the other hand, too much current can produce damage to the myocardium that may not be thermal in nature. Damage to the cell membrane and contractile structure may lead to depressed function, necrosis, persistent arrest, or single-cell fibrillation. Overdose of current may therefore result in failure to defibrillate. However, there clearly exists a therapeutic window for effective defibrillation. A uniform distribution of current throughout the heart ensures that no part of the heart is overdosed or underdosed, which contributes to the success of defibrillation.
Another factor contributing to success is the defibrillator waveform. The capacitor discharge waveform and, to a lesser extent, the AC waveform are less effective than the truncated exponential (trapezoid) or underdamped harmonic oscillator. Research related to implanted defibrillators has demonstrated new waveforms capable of greater effectiveness with less energy expended as well as greater therapeutic ratio. Biphasic waveforms are in use in Physio-Control (Redmond, WA), Philips Healthcare (Andover, MA), and rectilinear biphasic Zoll Medical devices.
The modern defibrillator works by temporarily storing defibrillation energy in a capacitor. A typical circuit for the underdamped harmonic oscillator waveform is shown in Figure 19-7 . The large capacitor is charged to the energy selected by the clinician and then discharged through the paddles. The standard formula for the energy (E) stored by a capacitor (C) at a voltage (V) is as follows: