Battery Status

Although not all systems provide the information in the same manner (some are graphic and others are numeric), data concerning battery function often include a measure of battery voltage, impedance, and current drain.^7–11 All three parameters are interrelated. The battery current drain is a measure of the average current being drawn from the battery, assuming 100% pacing at the programmed rate and output settings. Inhibition or pacing at higher rates, as might occur with intrinsic rhythm or sensor drive, respectively, directly affects the functional battery current drain, as do changes in lead or stimulation impedance. The reported current drain is only an estimate of the actual battery current drain. Newer devices are able to estimate remaining longevity based on true current drain measurements, giving a much more accurate assessment (see Fig. 29-1, C).

The measured current drain of the battery at a known set of parameters can be used to qualitatively assess the effect of programming changes on the longevity of the system. After the rate or output (or both) has been changed, the measured data can be reassessed and the effect of the programming change on longevity estimated. A simple calculation may be used to obtain an estimate of battery longevity when the pacemaker is relatively new. Most manufacturers provide the usable battery capacity reported in ampere-hours (A-hr). For this calculation, the battery’s capacity is converted to microampere-hours (µA-hr) and then divided by the measured battery current drain. The result is the anticipated number of hours of normal function at the programmed rate and output. This number, divided by 8760 (the number of hours in 1 year), yields an estimate of the longevity of the system in years. This calculation is best performed with a new system, before there has been significant battery depletion. Many newer pacemakers perform these calculations and display the remaining longevity in years or months, providing a relatively good estimate of remaining longevity.

Virtually every pacemaker incorporates a change in either the magnet or the demand rate (and sometimes in mode or other functions) to signal a level of battery depletion that warrants pulse generator replacement. Most of these changes occur when the battery voltage reaches a predefined level, which is specific for each manufacturer and even for each model of pacemaker. These changes are termed elective replacement time (ERT), recommended replacement time (RRT), or elective replacement indicator (ERI), with further changes in rate or function associated with continued battery depletion labeled either end of life (EOL) or end of service (EOS). Although these changes may occur abruptly, usually a 3- to 6-month period exists between activation of the ERT indicator and erratic behavior or device nonfunction caused by continued battery depletion. A marker to determine when to increase the frequency of pacing system surveillance (either transtelephonically or in the office) is included in some systems and is typically a magnet rate intermediate between a fully functional battery and the ERT indicator. Alternatively, this indication may be given to the clinician during interrogation of the device by a message on the programmer screen or on the programmer report. This is termed an intensified follow-up indicator and provides an indication of changing battery status before the ERT is reached.

Given the degree of programmability of most current pacemakers, a dual-chamber pacemaker programmed to pace all the time at maximal output on each channel may last only 2 years. An identical model programmed to outputs of 2.5 V or lower on each channel that are inhibited a significant portion of the time may be anticipated to function properly for 10 years or longer. It would be inappropriate to follow the second unit on a monthly basis beginning 1 year after implantation, whereas it would be equally inappropriate to plan only annual checks on a pacing system programmed with high outputs and 100% pacing. By following some measure of the status of the battery, either directly or indirectly, the clinician can achieve a qualitative assessment about when to increase the frequency of pacing system follow-up.¹²

There are two primary indicators of battery status: battery voltage and battery impedance. Battery voltage progressively decreases over time. The lithium/iodine (Li/I₂) power cell is the most common power source in pacemakers, whereas the power source in an ICD is lithium/silver vanadium oxide (Li/VSO). Some of the newer pacemakers are now using Li/VSO–polycarbon monofluoride (LVSO/P). The nominal unloaded output voltage of the Li/I₂ cell is 2.8 V, with each manufacturer triggering the RRT indicator at a specified battery voltage based on circuit design considerations. LSVO/P in pacemaker application has an initial voltage of 3.20, with an early drop to 2.90 V, where it plateaus, then drops to an elective replacement voltage of 2.60 V. The battery voltage and programmed output voltage are not identical. The “programmed output voltage” is achieved by modifying the 2.8-V battery output with either voltage multipliers or charge pumps to provide a higher voltage. The battery voltage can also be divided to deliver a voltage lower than 2.8 V. As energy is consumed, the battery voltage progressively decreases because of increased internal impedance. Even with decreasing battery voltage, the device circuitry is designed to maintain the stability of the programmed voltage delivered to the patient. The ability to track the progressive decrease in battery voltage and increase in battery impedance provides an effective guide to the clinician regarding the frequency with which the pacing system should be routinely checked for signs of battery depletion (Fig. 29-2). In addition, the measured battery voltage is affected by the battery current drain: the higher the current drain, the lower the battery voltage.

Figure 29-2 Pacemaker magnet rate, battery voltage, and battery impedance.

Graph over time demonstrates the inverse relationship between battery voltage and battery impedance. For this particular St. Jude Medical pacemaker, the magnet rate drops proportionally to the battery voltage. kOhm, killiohms; M, months; ppm, Pulses per minute; V, volts;

Battery impedance is inversely related to battery voltage. In the Li/I₂ power cell, the release of electrons is generated by the combination of the lithium ion with two iodine atoms. The result of this chemical reaction is the release of two electrons and the formation of lithium iodide. As it accumulates, the lithium iodide forms a resistive barrier between the anode and the cathode. With progressive cell depletion, the increasing amount of lithium iodide results in a rise in the internal impedance of the battery, which some pacemakers can measure and report in the telemetry data. The higher the battery impedance, the lower is the cell voltage.

Battery voltage and impedance may be tracked together or individually. If these measurements are provided to the manufacturer’s technical service engineers, along with the programmed parameters and an estimate of the percentage of pacing versus inhibition, the remaining longevity of that pacing system can be estimated with reasonable accuracy, assuming that the various parameters remain stable. These calculations have more recently been incorporated into the programmer interface to provide an online estimate of longevity in years or months. These estimates are sometimes shown as a range based on the tolerances of the battery measurements (i.e., giving minimum and maximum ranges of longevity), as well as an estimate based on present current drain and actual device use based on data acquired from the event counter diagnostics (see Fig. 29-1, B).

Lead Status

Although many systems provide data on lead function, including pulse voltage, current, charge, and energy, the most common measurement is lead impedance (stimulation resistance). Changes in lead impedance directly affect the other measures of lead function. The terms resistance and impedance, although technically different, are often used interchangeably by the clinical community. Impedance is a complex concept that reflects a changing environment involving a variety of factors. This results in fluctuations in the moment-to-moment resistance. The resistance to electron flow in a pacing system progressively rises during delivery of the stimulation pulse as a result of polarization at the electrode-myocardium interface and, as a continuously changing variable, it is appropriately termed impedance. Many pulse generators and pacing system analyzers make this measurement at a specific point within the pacing stimulus; this is a resistance. Other devices report the resistance as an average of the impedance throughout the pulse. The actual resistance to current flow imparted by the conductor coil is fixed and represents a small portion of the total stimulation resistance. The polarization at the electrode-myocardium interface, partly related to the surface area and geometry of the electrode, and the impedance associated with conduction of the pulse through the body’s tissues play a larger role in the overall resistance of the system. All these factors are incorporated in the single measurement termed lead impedance or, more accurately, stimulation impedance.

Stimulation impedance is affected by many factors, but primarily by the design of the electrode, including size, configuration, and materials. Manufacturers have designed some electrodes with high impedance values. For any given output, a high-impedance system reduces the overall current drain of the battery and effectively increases the longevity of the device. Other leads have been designed specifically for low polarization to allow for detection of capture with each pace stimulus. Polarization and impedance are not the same phenomenon, although one affects the other. For any given lead model, there may be a broad range of normal impedance values; for a specific lead within that model series, however, the impedance should fall within a relatively narrow range.

The clinician can use knowledge of lead impedance to monitor and identify a developing mechanical problem with the lead. This requires baseline and historical data to recognize subtle changes that may reflect conductor fracture or an insulation breach. It is essential to know what device is being used to make these measurements. As noted previously, different devices may obtain these data at different points of the pacing stimulus. Because of these differences, the impedance measurement obtained with a pacing system analyzer at implantation may be significantly different from that obtained by telemetry from the implanted pacemaker moments, if not years, later. This difference does not necessarily imply a problem. Furthermore, impedance may naturally evolve over time, with a fall in impedance occurring during the days to weeks after implantation, followed by a gradual rise toward the initial measurements. Multiple factors may affect impedance, particularly in a unipolar system. For example, measurements obtained during deep inspiration may significantly differ from those obtained during maximal exhalation. In the same patient, impedance measurements based on a single-output pulse have been reported to vary by 100 ohms (Ω), or even more, during the same follow-up evaluation, while remaining consistent with normal function. If a marked change in lead impedance from previous measurements (e.g., >200 Ω) is encountered during a routine follow-up evaluation, the pacing system should be further evaluated.^13,14 If the patient has no clinical symptoms and has stable capture and sensing thresholds, surgical intervention would be premature, although more frequent follow-up in the office or by remote interrogation would be prudent. However, a dramatic change in telemetered lead impedance in the presence of a clinical problem directs the physician toward the likely source of the difficulty (Fig. 29-3).

Figure 29-3 Measured data telemetry.

Rate-modulated dual-chamber pacemaker data report a low impedance on the atrial lead of 200 to 220 ohms (Ω) (Boston Scientific 1297). Although a unipolar lead, this impedance is very low and may be the result of an insulation failure. This also produces an increase in the battery or cell current, caused by a higher-output current, charge, and energy being delivered by the pacemaker (compare energy to that of the ventricular lead), whether or not it reaches the heart.

A dramatic fall in impedance usually reflects a break in the insulation.^15–17 In a unipolar system, an insulation problem provides an alternative pathway for current flow starting closer to the pulse generator, resulting in less energy reaching the heart. An outer insulation break in a bipolar lead tends to result in “unipolarization” of the lead. The amplitude of the stimulus artifact as recorded by an analog electrocardiogram (ECG) is determined by the distance the current travels in the tissues from the cathode (tip electrode) to the anode (ring electrode or housing of the pulse generator). A normally functioning bipolar pacing system in which both active electrodes are inside the heart, separated only 1 to 2 cm, results in a small stimulus artifact. In a unipolar system, the current travels from the tip electrode to the housing of the pulse generator, and therefore the stimulus artifact is large despite equivalent output settings. Some of the newer digital ECG designs result in a marked signal-to-signal variation in amplitude as a result of the digital sampling rate of the recording system. Other recording systems create an artificial uniform amplitude artifact with any high-frequency electrical transient, thereby precluding differentiation of a bipolar from a unipolar pacing system based on the analysis of the ECG recording.

In a previously stable system, a mechanical problem developing with the lead—either a breach in the insulation or a conductor fracture—results in a change in the stimulation impedance, which may be reflected by a change in the ECG-recorded stimulus artifact (Fig. 29-4). In a bipolar pacing system, an insulation defect between the proximal conductor and the tissues of the body is not likely to affect capture thresholds, but it results in a larger stimulus artifact, making it appear unipolar. Depending on the actual location of the insulation failure in either the bipolar or the unipolar lead, stimulation of the extracardiac muscle contiguous to the insulation defect may occur. Insulation failures may also attenuate the electrical signal reaching the pacemaker, possibly resulting in sensing failure.

Figure 29-4 Simultaneous electrocardiogram (ECG) rhythm strip and event markers.

Single-chamber, bipolar, rate-modulated pacemaker with an internal insulation failure of the lead. There are both oversensing (VS and VR) and functional undersensing when the true native complex coincides with the refractory period initiated by the oversensing. During the intervention for lead replacement, the measured lead impedance was 125 Ω. The event markers are both labeled (VP, ventricular pacing; VR, sensed event occurring during the refractory period; VS, ventricular sensing) and identified by different amplitudes of the marker artifacts. The surface ECG should also be examined. This recording system generates a large-amplitude “artifact” to simulate the paced output, with any high-frequency transient that is detected independent of the output configuration of the pulse. If no stimulus is detected, no artifact is generated. There are paced QRS complexes identified by the event markers that have no “pacing stimulus” on the ECG, whereas others have a large “unipolar” pacing stimulus visible. The identification of the pacing system malfunction was facilitated by the simultaneous event markers.

In a bipolar lead, an insulation defect developing between the proximal and distal conductors can present as a variety of clinical manifestations. Intermittent contact between the two conductors generates a voltage “transient” that the pacemaker can sense, resulting in inhibition or triggering, depending on the programmed sensing mode and the channel on which the problem has occurred. This small electrical signal is not seen on the surface ECG and is appropriately termed oversensing. If the two conductors remain in contact, current flowing down the distal conductor is short-circuited to the proximal conductor and never reaches the electrodes. This may result in a loss of capture and a further reduction in amplitude of the normally diminutive bipolar stimulus artifact on the ECG. In addition, loss of appropriate sensing may occur.

Programming of the output configuration to unipolar (most pacing systems, but not ICDs, now have this capability) may prevent the loss of capture and possibly undersensing from an internal short circuit. This does not prevent the oversensing behavior associated with make-break electrical transients generated by intermittent contact between the two conductors, as is caused by failed inner insulation. Typically, a bipolar lead with failed internal insulation exhibits a higher impedance when pacing is programmed to the unipolar rather than the bipolar output configuration.^18,19 This is opposite to the expected higher impedance of a bipolar system. Although programming to a unipolar configuration may result in an apparent resolution of a malfunction, it is not a cure and should be considered only as a temporary measure, one that allows the observed problem to be managed on an elective rather than emergent basis. The malfunctioning lead should be replaced expeditiously.

An increase in lead impedance may be the result of a conductor fracture or a connector problem.²⁰ When this occurs, the lead impedance often rises to high levels. It is inappropriate, however, to assume that a normal lead impedance is 500 Ω. Leads are available that are designed to have high impedance, with values ranging from 1500 to 2500 Ω. Other leads, at implantation, may have an impedance in the range of 250 to 300 Ω. Therefore, it is essential to look for a trend in serial lead impedance measurements rather than a single measurement. Impedance, taken in conjunction with the stability or changes in capture and sensing thresholds, is the best way to determine the condition of a pacing lead. A mechanical problem with the lead—either a conductor fracture resulting in high impedance or an insulation failure resulting in low impedance—eventually results in an overt clinical problem that can be identified by telemetric measurement of the stimulation impedance. If the impedance is sufficiently high, there is no current flow and no effective output. It must be understood that, although the telemetered event markers indicate a pace output, there may be loss of capture. The reduced current flow also results in a fall in the measured current drain of the battery. Note that any problem may be intermittent. This typically occurs when the two broken ends make contact at times but are separated at other times, or in the case of an insulation failure, when lead movement either opens the compromised area or pushes the edges of the break together, resulting in intermittent normal function.

Before the availability of telemetry for lead impedance measurements, physicians could obtain similar information by oscilloscopic analysis of the pacemaker stimulus. In the 1970s, some manufacturers included a photograph of the pulse wave contour at a variety of different impedance values with the technical manual accompanying each pacemaker. Some computer-based pacemaker ECG follow-up systems are able to record and display the pacemaker pulse at an expanded scale. This allows the pace waveform to be recorded and tracked as a routine part of the periodic pacing system evaluation. Although this feature is available, it is not routinely used, given the ready access to lead impedance measurements provided by the implanted pacing system.

Lead impedance measurements have an indeterminate incidence of false-negative results when the lead problem is only intermittently manifested. The system forces the pacemaker to function asynchronously in an effort to obtain the measurements. The measurements, however, are made over a few cycles. If there is a true lead problem, but it is intermittent and was not present when the measurements were being made, the telemetered lead impedance may be normal. Therefore, if a problem is suspected, repeated measurements should be performed. Use of provocative maneuvers, such as placing manual pressure along the subcutaneous course of the lead, or having the patient raise or in other ways manipulate the ipsilateral arm, may be required to unmask a problem.

Some pacemakers are able to report lead impedance measurements on a beat-by-beat basis, allowing the clinician to observe the digital readout of lead impedance on the programmer screen over many cycles. As with the oscilloscopic monitoring of the pulse wave morphology, this technique reduces but does not eliminate the incidence of false-negative results. A further enhancement is the “automatic periodic monitoring” of lead impedance, which results in a graphic display reporting the history of lead impedance performance. Measurements can be obtained continuously or daily. This increases the frequency of measurement, increasing the likelihood of diagnosing an intermittent lead problem and possibly intervening on an automatic basis (Fig. 29-5). Some devices respond by automatically programming the pacemaker from bipolar polarity to unipolar in the presence of either a marked rise in impedance, reflecting a conductor fracture,²¹ or a marked drop in impedance, consistent with an insulation failure.

Figure 29-5 Pacemaker lead impedance and intrinsic signal amplitude.

Graphic display (top) and tabular summary (bottom) of serial automatic atrial and ventricular lead impedance measurements as well as intrinsic signal amplitude from an implanted Boston Scientific pacemaker. This event counter reports measurements made on a daily basis for the past 7 days and the mean weekly measurements before that time, with the potential for extending back a full year. These data were retrieved at a follow-up evaluation.

Event Marker Displays

A variety of different displays have been used over the years. Although not intended for this purpose, the simplest event marker for sensed events was achieved with triggered mode pacing, particularly if the output configuration was unipolar. A sensed event would be marked by the simultaneous release of a pacing stimulus coincident with sensing. This was easily recorded on the ECG, because the stimulus artifact distorted the morphology of the native sensed complex. The next evolutionary step in this technology involved the pacemaker emitting a series of subthreshold stimuli of varying amplitudes to represent the release of an output pulse, a sensed event, and the end of a refractory period. Both these systems, the triggered mode and the series of markers, were limited by the signals requiring an intact lead system, to allow the pulse to be delivered to the heart through the lead and subsequently recorded for display by the ECG. With a bipolar output pulse, the small artifact might not be readily visible, particularly in the triggered mode, because it would be obscured by the larger native complexes. In addition, if there were an open circuit (most often a conductor fracture) or an internal insulation failure involving a bipolar lead, no marker would be visible.

The next improvement was to transmit coded signals representing paced and sensed events from the pacemaker to the programmer. The programmer reconstituted these signals into a series of varying-amplitude pulses or alphanumeric labels representing paced or sensed events. These could be either displayed on a monitor screen or recorded using a printer integral to the programming system, which allowed for the simultaneous recording of a surface ECG acquired by way of a separate set of cables. The simultaneous ECG and markers allowed the clinician to correlate the behavior of the pacemaker directly with the patient’s rhythm, to determine whether the system is functioning properly. A calibration signal composed of a series of different amplitude pulses or a legend of the cryptic labels allowed the clinician to interpret the various markers (Fig. 29-7). In an early series of markers, the largest pulse represented a pacing output, the intermediate one indicated a sensed event, and the smallest represented either the end of the refractory period or a sensed complex that occurred during the refractory period. With the advent of dual-chamber pacing systems, a marker pulse extending above the baseline identified atrial events, and ventricular events were labeled with a pulse extending below the baseline. This type of marker system has become obsolete, and there are very few devices remaining as active implants that use this scheme.

Figure 29-7 Example of formerly used event marker annotation.

Lines of various amplitude are used to designate paced, sensed, and refractory events. Lines above the baseline are used to describe atrial events, and lines below the baseline to describe ventricular events. This type of annotation has been made obsolete by the current annotation systems.

A further refinement to this system was the addition of alphanumeric annotations, which made the simple (yet cryptic) system just described obsolete. Two common sets of notations have been used and are summarized later. These are not the only possibilities, because some manufacturers use a unique set of symbols to identify paced and sensed events, with the location of the symbol on the recording identifying whether it represents an atrial or a ventricular event. Others have expanded on the set of labels and provide, on command, a detailed explanation.

One method for single-chamber pacing systems with event marker telemetry uses the letter P to reflect a paced event and S to represent a sensed event. This may cause confusion with other dual-chamber systems in which the letter P reflects a sensed atrial event indicative of a native P wave (with which most medical personnel are familiar, based on their knowledge of the standard ECG). The interpretation of the alphanumeric event markers is often product specific, and the clinic personnel who evaluate the pacing system must take this into account (Fig. 29-8).

Figure 29-8 ECG and EGM tracings with event marker legend.

Lower ECG rhythm strip (top) with simultaneously recorded atrial (AEGM, middle) and ventricular (bottom) electrograms. Labels across bottom identify specific events reflecting the behavior of this Boston Scientific pacemaker. In these tracings, the atrial refractory sensed events (AS) are caused by far-field R-wave sensing that is also visible on the AEGM channel.

In the case of dual-chamber pacing or when the pacemaker knows that it is providing atrial pacing and sensing, a native atrial depolarization may be identified as either the letter P or the combination AS. P is taken from the standard ECG identifier for an atrial depolarization. AS is one abbreviation for an atrial sensed event. An atrial stimulus may be identified by either the letter A or the combination AP for atrial paced event.

Analogous lettering is used on the ventricular channel. Again, single-chamber pacing in some systems might still use the letters P to represent a paced event and S to refer to a sensed event. Other systems use R to refer to a native ventricular depolarization, which is generically called an R wave on the standard ECG. This may also be identified by the letter combination VS for a ventricular sensed event. The letter V in one system might be used to represent a ventricular pacing stimulus, whereas the identical event may be labeled VP (for ventricular paced event) in other systems.

Other symbols may be used for a paced or sensed event and are usually displayed with an identifying key on the resultant printout. Each paced or sensed event may also be identified by a vertical line, going up in the case of atrial events and down for ventricular events. In some cases, a difference in the amplitude of these lines has been retained in conjunction with the alphanumeric lettering. Indeed, as the complexity of the devices increases, the variety of symbols and letters identifying specific events and behaviors is also increasing (see Fig. 29-8).

In many systems, interval measurements between the various events are either calculated automatically and displayed, or measured after cursors are aligned with specific events. There may be a time delay between the actual sensed or paced event and the release of the event marker. This is caused by the finite period that is required for the pacemaker to recognize the event and then transmit the appropriate information to the programmer through the telemetry channel. In most cases, priority is given to the normal function of the pacemaker, with the diagnostic marker feature being delayed. Differences of up to 40 msec between the actual event and the resultant marker have been noted.

Laddergramming

Laddergramming is an advanced adjunct to the interpretation of clinical arrhythmias. It was incorporated in some programming systems by using the known programmed parameters of the pacemaker combined with the event marker information telemetry from the implanted pacemaker^23–26 (Fig. 29-9). For all the elegance of these graphic interpretations of the pacemaker’s behavior, they still require the simultaneously recorded ECG rhythms. Although the pacemaker may be functioning normally, in that it is behaving properly with respect to its programmed parameters, failure to capture or properly sense would not be diagnosed from the various diagrams and markers without the simultaneously recorded surface ECG.²⁷ Laddergramming has become a teaching tool at this time, because no programmers currently provide this level of graphic description.

Figure 29-9 Laddergram from rate-modulated dual-chamber pacing system.

The event markers are identified by a variety of symbols displayed in a key to the left of the tracing. In addition, battery (cell) impedance and projected longevity are reported in the upper right portion of the tracing, and selected programmed parameters are shown in the upper left quadrant. AV, Atrioventricular; RRF, rate response factor.

Data from a META DDDR Model 1250 Telectronics Pacing Systems, now Pacesetter, St. Jude Medical CRMD, Sylmar, Calif.), displayed with a Model 9600 programmer.

Event Marker Limitations

Several limitations are associated with event marker telemetry. First, used alone, markers report the behavior of the pacemaker but do not allow the clinician to determine whether this behavior is appropriate for the patient.²⁷ In this regard, markers are analogous to the hand-held digital counters that report pacing rates and intervals. The counters only detect and report the pacing stimuli, not whether these output pulses are effective in causing a cardiac depolarization, or whether native events are properly and consistently sensed. A long interval between consecutive pacing stimuli could reflect normal function because the pacemaker responded appropriately to a native complex, or it could represent a system malfunction with oversensing or no output (e.g., from intermittent lead fracture). Neither event marker telemetry nor the digital counters should be used independently of a simultaneously monitored ECG rhythm.

The second limitation is that the report of a stimulus output does not mean that there was appropriate capture. Confirmation of capture requires an ECG or concomitant hemodynamic pulse monitoring. In addition, the markers may report that the pacemaker is sensing a signal, but if this signal is not visible on the surface ECG, there is cause for concern. Is it an appropriate signal that the pacemaker should be sensing (some intrinsic atrial rhythms may not be visible on the specific lead that is being monitored), or is the system responding to an inappropriate or nonphysiologic signal, such as the make-break potentials associated with a breach of the internal insulation in a coaxial bipolar lead?

Most pacemaker event marker telemetry is limited to real-time recordings. The markers must be telemetered from the pacemaker to the programmer or another display system while the rhythm is being actively monitored. Neither the pacemaker nor the programmer can retrospectively provide markers for a previously recorded rhythm. If the pacemaker is responding to events that are not visible on the surface ECG, the event marker simply confirms this fact but does not identify the specific signal. An evaluation of sensed but otherwise invisible events requires intracardiac electrogram telemetry or an invasive procedure to record the signal from the implanted lead. Many premium-level pacemakers can store event markers with or without electrograms. Markers may be stored using a patient trigger (e.g., magnet application),²⁸ a simple battery-operated radiofrequency transmitter, or on spontaneous activation of predefined algorithms or sequences of sensed or paced events.

Limitations of Electrogram Telemetry

The apparently adequate amplitude of the telemetered EGM for sensing does not mean that the pacemaker will sense it. The telemetry amplifier may use filters different from those in the sensing circuit, providing qualitative rather than quantitative data on the EGM (see Fig. 29-11). In most pacing systems, telemetered EGMs also have limitations similar to those of event markers. EGMs are real-time recordings and cannot be retrospectively provided by the programmer to facilitate the interpretation of an earlier ECG.

Real-time telemetry allows the physician to analyze the behavior of the pacing system when the patient is in the physician’s office or clinic while these diagnostics are being accessed with the programmer. This is impractical over a long period. In addition, it does not allow the patient to move around much while these recordings are being obtained. Long-term monitoring of pacing system behavior requires a Holter monitor, a loop memory recorder, or event counter telemetry, depending on the desired degree of precision. Pacemakers with microprocessors and significant random access memory (RAM) can now store select EGMs,^47,48 as well as event markers, when triggered by the patient or by a predefined set of circumstances. This may greatly reduce the need for Holter or other monitoring techniques to evaluate intermittent symptoms in patients with pacemakers.

Total System Performance Counters

The simplest systems keep track of the number of times a pacing stimulus is released or a native complex is sensed. This allows for an assessment of the degree to which the pacemaker was used by calculating the percent pacing in each chamber. A refinement of this ability allows the pacemaker to diagnose bradycardia and, in the dual-chamber modes, to determine whether the bradycardia was the result of sinus node dysfunction or AV block. Event counters with this ability have also been termed diagnostic data, implanted Holter systems, and data logging.^49–55

With the introduction of dual-chamber pacing (specifically the DDD mode), the potential interactions between the pacemaker and the patient increased from two (pacing or sensing in one chamber) to five (pacing or sensing in the atria or ventricles and sensing ventricular activity without preexisting atrial activity). The situation became even more complex with the introduction of multiple-site ventricular and atrial pacing configurations. Knowledge of how the pacing system has behaved over time, combined with the clinician’s knowledge of the patient, provides invaluable information toward assessing the overall performance of the implanted system.

The various annotations described in the event marker section were combined to provide a cryptic description of the different operational states of the DDD mode. These include an atrial sensed event followed by a ventricular sensed event (AS-VS or PR), indicating that the pacemaker is inhibited on both channels. Atrial sensing followed by a ventricular paced event (AS-VP or PV) refers to P-wave synchronous ventricular pacing. Atrial pacing with intact AV nodal conduction or functional single-chamber atrial pacing is indicated by AP-VS or AR. Base-rate pacing in both chambers is represented by AP-VP or AV. A ventricular sensed event not preceded by atrial activity, either paced or sensed, is a premature ventricular event (PVE), identified by the letter R or the combination VS. Some systems label these events premature ventricular contractions, or complexes (PVCs).

This information may be collected in conjunction with events occurring in specific rate bins, allowing for a detailed report of heart rate distributions. Additional data collected include the percentage of pacing in the atrium and ventricle, length of time at the maximum tracking rate, and the number of episodes in which the pacing system reached the programmed upper rate limit (Fig. 29-12). Additional data may be recorded regarding mode-switching events, pacemaker-mediated tachycardia (PMT) termination algorithm use, or the number of times any one of several other special features was activated. The counters are able to continue to accumulate data until one of the pacing state bins is full and cannot accept additional information. At this point, one or more of the counters are frozen. The volume of data that can be stored depends on the memory capacity of the pacemaker that is dedicated to these features.

Figure 29-12 Heart rate, event, and mode switch histograms.

Data represent the performance of the pacing system over the preceding 448 days, with monitoring of every event. Patient had a St. Jude Medical device programmed to DDDR mode. The event histogram is a graphic display of the relative percentage of events within each of the five basic pacing states, whereas the heart rate histogram is a graphic display of the rate distribution based on atrial sensed or atrial paced events. This distribution reflects normal chronotropic function of the sinus node in this patient, whose pacemaker was placed for intermittent atrioventricular block (almost all the atrial paced events are at the base rate). The percentage of time per rate bin represents a normalized display of the relative percentage of atrial paced or atrial sensed events and premature ventricular sensed events in each rate bin during this period. The AMS summary and V Rates During AMS indicate the atrial rates during mode switch, the duration of the mode switch events, and the ventricular response during the mode switch events. These data can be used to manage antiarrhythmia medication as well as AV node–slowing medications.

The following example illustrates the amount of data that can be stored, or the quantity of memory. All data are stored in a binary code, represented by a series of 1s and 0s. A binary unit of memory is called a bit. Eight bits are a byte and result in 256 combinations of 1s and 0s. If each series of combinations represents a single item of data, a total of 256 pieces of information can be stored in a system with eight bits of memory. As the number of bits increases, the amount of data that can be stored increases exponentially. If each pacing state is represented by a single combination of bits, and there are 24 bits in the RAM devoted to these counters, more than 17 million events can be counted and stored in memory for retrieval at the time of routine pacing system follow-up. Reading the counters requires access to the memory banks in the pacemaker through the use of specific coded commands from the programmer. One DDD system introduced in the mid-1980s had sufficient memory to provide a summary of pacing system behavior for a period of about 6 months. As memory capacity has increased and data compression algorithms have evolved, these systems are now able to store much more data for periods longer than 1 year.

Storing a simple marker of the pacing state and rate requires relatively little memory compared with storing waveform data representing the rhythm, as depicted by a consecutive series of EGMs. A theoretical 100-Hz bandwidth with 200 samples per cardiac cycle at eight bits of resolution per sample, recording the entire rhythm during a 24-hour period, would require 140,000,000 bits/day, even if the heart rate was a steady 60 beats per minute (bpm).⁴⁷ Most implantable defibrillator systems store a series of EGMs preceding and following the delivery of antitachycardia therapy. Premium pacing systems use certain triggers to initiate EGM storage. These triggers may be a high atrial or a high ventricular rate, initiation of mode switching, or activation of any of several other specialized algorithms. Storage may also be triggered by application of the magnet or use of a simple transmitter device by the patient when symptoms are present. An early method to save memory was to provide “snapshots” of representative complexes; the trigger to store these data in an ICD was delivery of antitachycardia therapy. Extensive storage of cardiac rhythm data within pacemakers became possible with increasingly sophisticated data compression algorithms and more memory than previously available. Higher-bandwidth data transmission channels are required for this volume of data storage, particularly for long series of rhythms, so that transmission can be accomplished as efficiently and as quickly as possible.

Implanted looping memory event recorder (implantable loop recorder, ILR) capability, complete with stored rhythms, is now available as a stand-alone device, in ICDs, and in high-end pacemakers. Most of the older pacemakers, however, continue to store only data reflecting the total number of times a pacing state was encountered and, in some systems, the distribution of rates or intervals, rate ranges, or mean rates within these pacing states. This information provides the physician with an overview of the function of the system in the patient. Predominant base-rate pacing is expected in patients with marked sinus node dysfunction and in those receiving certain medications such as beta-adrenergic blockers or some calcium channel blockers. On the other hand, predominant base-rate pacing (AP-VP) in patients whose primary indication for pacing was complete heart block suggests either the development of sinus node dysfunction or primary atrial undersensing. In these same patients with high-grade AV block, frequent counts of AS-VS and PVEs, particularly in the absence of known ventricular ectopy, suggests episodes of ventricular oversensing or resolution of the AV block. Large numbers of counts regarding PMTs warrant a reassessment of retrograde conduction and possible adjustment of the programmed PVARP or PMT prevention and termination algorithm. It may also mean only that the maximum tracking rate (MTR) is being reached, as some devices use the MTR as the trigger for PMT intervention. A large number of counts at the MTR suggests that the upper rate limit should be reassessed, because it may be too low for the patient’s physiologic needs. It might also indicate pathologic atrial rhythms. Event counter telemetry can help reveal the overall function of the pacing system; a variation from the clinician’s knowledge of the patient may suggest the need for a further evaluation.

When the additional ability to monitor a series of rates and pacing states is added, chronotropic function can be assessed by the distribution of atrial sensed rates (AS-VS or PR and AS-VP or PV; see Fig. 29-12). Furthermore, atrial pacing at rates above the programmed base rate reflects sensor drive in the DDDR pacing systems, providing an overview of the sensor behavior. The ability to report both rates and pacing states can provide the clinician with a better insight into the cause of PVEs as well as overall pacing system function. True premature ventricular contractions tend to occur at short coupling intervals, which is equivalent to a rapid rate. Large numbers of PVEs might suggest recurrent ventricular arrhythmias.^56,57 If there are large numbers of PVE counts at relatively low rates, the clinician should consider episodes of atrial undersensing or accelerated junctional rhythms, which would also fulfill the pacemaker’s criteria for a PVE (i.e., a sensed R wave not preceded by an atrial event).

Subsystem Performance Counters

An increasing number of specific algorithms and capabilities are used either intermittently or potentially independently of the functional performance of the pacing system. An early counter representative of this capability is the sensor-indicated rate histogram, which reports the distribution of pacing rates that would have occurred had the sensor been totally controlling the pacing rates.^58,59 This counter reports the sensor-defined rates, even if the pacemaker was inhibited by a faster native rhythm or the actual functional rate was being controlled by the sensed atrial activity (Fig. 29-13).

Figure 29-13 Sensor-indicated rate histograms showing heart rates for two patients.

The bars show the actual heart rates and are dark if atrial paced or white if atrial sensed (St. Jude Medical pacemaker). The open circles indicate what the activity sensor would have driven the heart rate to if the sensor-indicated rate was faster than the patient’s intrinsic heart rate. A, This histogram is from a patient with a normal sinoatrial (SA) node, showing that the patient’s heart rate exceeded the sensor-indicated rate in all bins except the base rate bin. B, This histogram is from a patient with sinus node incompetence; thus almost all the atrial events are paced, and at the sensor-indicated rate.

Other subsystem performance counters include automatic mode switch histograms, reports of the cumulative length of time for which the system functioned at MTR, number of times the PMT algorithm was enabled, high atrial or ventricular rate episodes (Fig. 29-14), and sudden rate-drop episodes. These diagnostic counters provide detailed information about the use of a specific algorithm or system behavior that is not activated daily or is not visible on the ECG, making identification difficult using standard recording techniques such as a Holter monitor.

Figure 29-14 High-rate episodes on AEGM.

High-rate episode summary (top) and high-rate episode graphic with stored atrial electrogram (AEGM; bottom). These data facilitate the clinician’s ability to assess the behavior of the pacing system between office visits. In this case, the specific episode associated with the AEGM reported atrial paced events occurring above the trigger level for this event counter, which was also above the maximum sensor rate. The electrogram demonstrates atrial pacing with a large, far-field R-paced ventricular event; presumably the labeled high-rate episode was caused by far-field sensing. This knowledge may be helpful in further programming of the pacemaker.

Data from Medtronic Thera DR Model 7940.

The principal limitation associated with the two types of counters so far described is that counts are placed in a bin, either the pacing state alone or the pacing state and rate, which provides a one-dimensional view of the system’s behavior. If the period of monitoring is short, as during a casual or brisk walk, the clinician can reasonably assess the behavior of the pacing system. Longer periods of monitoring accumulate and overlap the results of many activities. This precludes an assessment of the pacing system’s behavior during a specific activity, or the identification of a symptomatic episode occurring at an earlier time. The ability of the system to store pacing state and rate data with respect to time (i.e., time-based event counters) has been variably termed event record or pacemaker Holter systems.^60–62 Technically, this is not yet a true Holter monitor, because continuous rhythm recordings are not retained in memory, even though rate data may be available.

Time-Based System Performance Counters

Time-based event counters require an extensive amount of memory. Not only are the pacing state and rate stored, but these data must be stored with respect to preceding and subsequent events. The data cannot be simply dumped into a rate or pacing state bin. Each piece of data (i.e., every cardiac event) must have a time reference, which takes significantly more memory than simple histogram-type counters.

Two techniques are available for storing real-time data. One is to continue to accumulate the data until the memory is full and then to freeze the recorded events. Freezing all the counters at the same time maintains the ratios between the events. A clearing function is usually provided to reset the counters. In some devices, reprogramming of any parameter or specific parameters (e.g., rate or pacing mode) can result in the automatic clearing of these data. This “initial events” or “frozen” recording technique is especially useful for short-term monitoring, as in office-based exercise evaluations. It also allows the system’s response to exercise to be evaluated, often without the need for simultaneous ECG monitoring.

The other option is to collect the data continuously. As the counters fill, new data are added at the expense of the oldest data (Fig. 29-15). This has been called rolling trend, final trend, or continuous data storage and is based on the “first in, first out” (FIFO) principle. When the patient is seen in follow-up, the data acquired over the time immediately preceding the interrogation of the system are available for review. The patient can often recall symptoms and activities during this period. These can be correlated with the behavior of the pacing system at the time of the symptoms. In this manner, the clinical staff caring for the patient may further assess chronotropic function, the behavior of the pacemaker during a variety of special or usual activities, and whether the sensor and other algorithms are behaving appropriately. The clinician may gain insight into the cause of palpitations and other symptoms that may have occurred during this time by correlation of the recorded data and the patient’s complaints or activities.^51,58–62

Figure 29-15 Event record display.

The sampling interval was 26 seconds, which provides a detailed overview of the behavior of the pacing system for the preceding 60 hours. The top display is an overview of the moment-to-moment behavior of the pacing system, with the vertical lines representing the maximum and minimum rates during each time period (the total displayed interval is divided into 40 equal time segments) and with the cross-bar representing the mean heart rate. The vertical line is displayed if there are two or more complexes in the specific time bin. The display scale can be expanded to display individual events, in which case a specific alphanumeric label is shown, consistent with the legend in the upper left. The same data may be displayed as an event bar graph (equivalent to the event histogram) or as a rate bar graph (equivalent to the heart rate histogram). The solid bars represent atrial paced events; thus atrial rates greater than the programmed base rate represent sensor drive. Atrial rates lower than the programmed base rate represent “sleep mode” behavior that was also programmed. The time graphs automatically divide the overall monitoring period into six equal segments.

Display is from a Pacesetter Trilogy DR+ Model 2364 (St. Jude Medical).

Sensor-input data have been combined with the actual rates achieved to facilitate reprogramming of the sensor parameters. Once acquired, the data from an exercise session are retained in the memory of the programmer, uploaded to the programmer, and then displayed on a screen. The system’s performance is based on a given set of sensor parameters that were in effect at that time. Because the pacemaker has the actual sensor-signal data, it can calculate how the unit would perform in response to a different set of sensor parameters. If a new set of sensor parameters is entered into the programmer, the curves that were based on the original input data are redrawn to display the system’s projected behavior in response to the new set of parameters. The clinical staff can then determine the best sensor parameters for the individual patient without having the patient perform repeated activities (Fig. 29-16). This feature has been termed redraw,⁶³ exercise test, or prediction model.⁶⁴

Figure 29-16 Exercise test report.

The lower line on the graph represents the actual behavior of the pacing system during a hall walk with the settings as noted to the right of the graph as “tested.” Patient has St. Jude Verity pacemaker. The thicker line above represents the behavior of the pacemaker under sensor drive at the proposed or “modeled” values. By using this type of feature, the patient only needs to take a single walk to set up the activity response of the device.

Limitations of Event Counter Telemetry

The major value of event counters is providing a review of system performance over time. This is not available from the real-time diagnostic features of measured data and event counter and EGM telemetry. The data, however, are interpretable only after a detailed assessment of the capture and sensing thresholds, in a manner analogous to event marker telemetry. The pacemaker can only store information that it knows about: output and sensed events, sensor data, and activation of unique algorithms. If there is a problem with undersensing, the pacemaker releases a pacing pulse, and the counters report a paced rhythm. Similarly, a large number of sensed events may cause the pacemaker to be inhibited on the respective channel. The counter simply reports a large percentage of sensed events, but this may be clinically inappropriate if the cause is oversensing.

With respect to pacing, a rhythm that is predominantly composed of AV- or PV-paced complexes does not necessarily indicate complete heart block. It might result from a programmed AV delay that was not sufficiently long to allow for full conduction through the AV node. Although the ventricular complex could have been fully paced, it could also have been a fusion or pseudofusion beat. The counters cannot differentiate between these complexes, because the native complex had not yet been sensed, the timer had been completed, and an output was delivered. Another situation would be a total loss of ventricular capture with intact AV nodal conduction (Fig. 29-17). The first conducted R wave in this figure would not have been sensed, because it occurred during the refractory period that followed the ventricular output pulse. An identical result in the counters would occur with intact AV nodal conduction but with ventricular undersensing.

Figure 29-17 Telemetry data limitations.

Although this printout was obtained during a ventricular capture threshold test, it demonstrates the limitations of some of the telemetry data in the absence of a simultaneously recorded electrocardiogram (ECG). The last two ventricular output pulses were subthreshold, resulting in a loss of capture. The endogenous P wave conducted, but with a marked first-degree atrioventricular (AV) block, placing the R wave outside the ventricular refractory period initiated by the subthreshold ventricular output, allowing it to be sensed. Based on event marker and event counter data, this patient would be interpreted as having frequent premature ventricular events (PVEs) interspersed in a stable atrial sensed, ventricular paced rhythm; however, in actuality, there was a loss of ventricular capture with intact AV nodal conduction. Key programmed and recording parameters are also included with each printout to facilitate interpretation; EGM, Electrogram.

Data from Pacesetter Synchrony II Model 2022 dual-chamber pacing system (St. Jude Medical CRMD, Sylmar, Calif.)

The event counters simply report the system’s performance without making any statement or judgment about the appropriateness of this behavior. The information provided by the event counters can be interpreted only after the programmed parameters are known, the pacing and sensing thresholds have been determined, and the status of the native rhythm is known. If the pacemaker is programmed to provide a good margin of safety for both pacing and sensing, the event counters are likely to be an accurate reflection of the patient’s clinical rhythms. These data also provide insight into the degree to which the patient requires pacing support at the current programmed parameters, the chronotropic function of the sinus node, and responsiveness of the sensor. Evidence of oversensing or lead dysfunction renders the event counter data suspect with regard to these findings, although a marked change in event counter data can provide the clinician with a clue to a developing problem.

The event counter data cannot be adequately interpreted without knowledge of how these data are collected, the clinical status of the patient, and any unique behavioral performance of the implanted pacemaker. This is illustrated by Figure 29-14, which is an atrial high-rate graphic from a Medtronic Thera DR pacemaker. The reported rates were greater than 200 bpm, yet the graphic suggests that these were atrial paced events and that the device is not capable of being programmed to rates this high. It is essential to know how the manufacturer calculates rates. In this case, the device measures the interval between atrial events; if the event preceding the atrial output occurred during the refractory period (labeled AR by the markers), the reported atrial paced rate would be based on the AR-AP interval. With this knowledge, the graphic report can be understood. If this fact were not known, the graphic would suggest a pacing system malfunction.

Atrial fibrillation (AF) is a major clinical problem that is common in patients with cardiac implantable electronic devices (CIEDs). Management can be improved by knowing whether AF is present, the frequency of AF, and the duration of AF episodes. Knowing these data allows the clinician to determine the need for antiarrhythmia therapy, for ventricular rate–slowing medication or interventions, or for anticoagulation. Several diagnostic counters are available to assist in the management of AF (Fig. 29-18). These may consist of the number and duration of the episodes, atrial rates and ventricular rates during the episodes, and specific dates and times so that patient symptoms can be correlated to events.

Figure 29-18 Event log and histogram.

A, Event log from a Medtronic pacemaker with mode switch algorithm showing date, time, and duration of the atrial arrhythmia, presumably atrial fibrillation (AF). Maximum atrial and ventricular rates are also provided. B, Mode switch histogram from a St. Jude pacemaker showing the number of episodes, peak atrial rates, and the duration of the AF episodes. AF burden graph is also shown for a St. Jude pacemaker as well as AF trend for a Medtronic pacemaker. This can assist in the management of patients regarding need for anticoagulation or antiarrhythmia drug therapy. C, Cardiac Compass Report from a Medtronic device implanted into a patient in January 2005. The patient went into persistent AF 2 weeks after implantation. Note the presence of AF 24 hr/day on the top graph for most of the period monitored. The lower percentage of AF during June and July could have been caused by intermittent sinus rhythm or by AF undersensing. The bottom graph shows the average and maximum ventricular rates while the patient was in AF and can be used to determine whether further therapy to control the ventricular response is needed.

The one caution in terms of accepting data from these counters is the possibility of oversensing in the atrial channel, which would cause inappropriate assignment of events to these diagnostics. The most common cause of false information is sensing of the far-field QRS complex by the atrial lead (Fig. 29-19). This leads to double-counting of the heart rate and incorrect classification of the event as atrial high rate. Use of stored EGMs to verify the cause of the high-rate episode is advised whenever possible, to avoid inappropriate treatment for nonexistent AF.

Figure 29-19 Far-field QRS complex.

The top line displays the event marker telemetry, showing that the device is sensing both the intrinsic P wave as well as the far-field QRS complex on the atrial lead. The middle trace is the intraventricular electrogram, and the bottom trace is the intra-atrial electrogram (AEGM). Note that the QRS is easily seen on the AEGM, thus causing the pacemaker to double-count the atrial rate.

Failure of Output (No Artifact Present)

In the pacemaker-dependent patient, failure to pace represents a potentially catastrophic and lethal situation. The ECG appearance in a single-chamber system is that of no pacing artifact when the intrinsic rate is below the lower rate limit of the device. This may be intermittent or continuous. In a dual-chamber device, the clinician may see an atrial pace alone, ventricular pace alone, or no pacing artifacts at all. Of course, any of the latter may be entirely normal depending on the patient’s sinus rate, AV nodal conduction, A-V interval settings, enabling of hysteresis or one of its analogs such as sleep mode or rate-drop response, and sensor settings. Box 29-7 lists the causes of a no-output condition. In evaluating this situation, the clinician must be certain that a malfunction truly exists before taking any corrective action. Remember that the ECG (especially as visualized in a single surface lead) may not display the pacing artifact because of technical factors (Fig. 29-26). This is particularly common in bipolar systems and systems programmed to a lower-than-nominal voltage output. It is important to recognize that the problem may be failure to capture rather than failure to output. Access to telemetered event markers and measured data facilitates the evaluation. Event markers identify the release of an output pulse or recycling resulting from a sensed event, even if this is not visualized on the ECG. If lead impedance is normal (with the previously noted caveat that the problem may be intermittent and multiple measurements should be obtained), this probably reflects a capture problem if there is an output, or an oversensing problem if there is a sensed event. A high impedance suggests an open circuit, whereas a low impedance suggests an insulation failure. Both are probable mechanical problems with the lead in a chronically implanted system. In a newly placed system or when the leads have been recently replaced into the pacemaker, this may be caused by malalignment of the lead in the connector, a loose set screw, or failed spring-loaded connector.

Figure 29-26 12-lead ECG of patient with normally functioning bipolar pacing system.

Note that a small pace artifact is seen in lead V4, whereas other leads do not show it at all.

If the markers fail to demonstrate an ineffective output or an inappropriate sensed event, normal function (e.g., hysteresis) should be considered. At this point, the programmed parameters should be double-checked to determine whether hysteresis or one of its analogs was programmed. Pacemakers have been returned to their manufacturers as defective when hysteresis was enabled. In these cases, the physician was either not aware of the programming or did not understand this feature.^112–114 Other causes of intermittent prolongation of the pacing interval are listed in Box 29-7.

Another source of evaluation error is failure to recognize intrinsic complexes that are present and are properly sensed, causing an appropriate device inhibition. This problem may occur in the hospital’s telemetry unit, where only a single lead is monitored. Premature beats may be present, with the ectopic complexes appearing nearly isoelectric (Fig. 29-27). Although the observer may not recognize the PVC or premature atrial contraction (PAC) on casual observation, the pacemaker senses these events and is appropriately inhibited. The use of multiple leads or a 12-lead ECG recording virtually eliminates this cause of pseudomalfunction.

Figure 29-27 ECG rhythm strip of pacemaker programmed to VVI with a demand rate of 50 bpm.

This appears to show an inappropriate bradycardia. Note the low-amplitude waves that map out to the escape interval. These are premature ventricular complexes/contractions (PVC) that appeared almost isoelectric in this lead.

The most common cause of lack of an output pulse in unipolar systems is oversensing. The pacemaker, for all its complexity, operates from a relatively simple set of rules. Any event that generates an electrical signal of adequate strength and frequency content (during the alert period) is sensed and resets the timing cycle. Oversensing is the sensing of a physiologically inappropriate or nonphysiologic signal. A cause of intracardiac oversensing in AAI systems is sensing of the far-field QRS complex¹¹⁶ (Fig. 29-28). This is most common in unipolar systems when the lead is placed close to the tricuspid valve, with high atrial sensitivity settings, and with short atrial refractory periods. In systems that can deliver antitachycardia pacing (ATP) therapy, oversensing may result in an inappropriate diagnosis of a tachycardia and therapy delivery.¹¹⁷ In dual-chamber systems, atrial far-field sensing is also one of the most common causes of inappropriate mode switching, because the pacemaker is seeing double the actual heart rate. Far-field sensing of the atrium is uncommon in VVI systems because of the diminutive size of the P wave, as seen electrically from the ventricle. Although reported, it is usually associated with other problems, such as an insulation abrasion on the atrial portion of the ventricular lead.¹⁵ Sensing of the intracardiac T wave may occur in VVI systems with similar results (Fig. 29-29), although this was a more common problem in the early days of pacing, as was oversensing of the polarization artifact generated by the output pulse. Other chapters in this text provide a description of polarization and the evoked response. Oversensing of this event occurred with short refractory periods combined with high outputs on certain lead systems. This has also been reported in ICD systems that require short refractory periods and broad band-pass filters in the sense amplifier.

Figure 29-28 Intracardiac electrogram (IEGM) recorded from unipolar atrial lead.

Note the large size of the far-field ventricular electrogram (VEGM), labeled R. If the near-field atrial electrogram (AEGM), labeled P, were not as large, the device could undersense the atrium, sensing the VEGM in error. ECG, Electrocardiogram.

Figure 29-29 ECG rhythm strip of a device programmed to VVI at 60 bpm.

T-wave oversensing is present. The last sensed event can always be determined by mapping back by the escape interval.

The most common source of a physiologic extracardiac signal that causes oversensing is myopotential inhibition. The pacemaker is capable of sensing muscle depolarizations—not only those of the heart, but also those of noncardiac structures such as the pectoralis, rectus abdominis, and diaphragmatic muscles. Myopotential inhibition of pacemaker output can be seen during arm movements (pectoral implantation) or when the patient sits up or does a Valsalva maneuver (abdominal implantation; Fig. 29-30). This occurs primarily in unipolar systems, because the anode, the pacemaker’s case, remains in proximity to these muscular structures.^{88,89,118–124} Nonphysiologic electrical signals may also cause oversensing. Formerly, electromagnetic interference (EMI) from microwave ovens was a significant problem when pacemakers were encased in plastic or epoxy, with the relatively large antenna created by the discrete components. Newer devices are shielded from EMI by their metal cases and filters. Integrated circuits provide for a smaller antenna, although strong EMI sources can still cause oversensing. Table 29-1 summarizes EMI sources and their potential effects.^125–142

Figure 29-30 A, ECG rhythm strip shows inappropriate inhibition caused by myopotential oversensing. B, ECG rhythm strip of the same pacing system, with a magnet placed over the pulse generator. If a loss of output were still evident with the magnet “on,” the problem likely would not be oversensing.

It has been well documented that intermittent metal-to-metal contact can create electrical signals. These make-break signals may be of sufficient amplitude to be detected by pacing systems. Signals exceeding 25 mV have been recorded in some cases (Fig. 29-31). These large signals present a major problem: the false signal cannot be managed by programming the device to a less sensitive setting, because the nonphysiologic signal is far greater than the physiologic atrial or ventricular depolarization. The sources of these electrical signals include a loose set screw, fracture of the conductor coil, failed insulation between the anode and cathode conductors in a bipolar system, interaction between active and abandoned pacing electrodes, and a loose retractable fixation screw.^142–150 Other mechanisms can also cause these signals, such as capacitance between the anode and cathode of a coaxial bipolar lead with failing insulation.

Figure 29-31 A and B, Ventricular intracardiac electrograms (IEGMs) as telemetered by a pulse generator, recorded simultaneously with a surface electrocardiogram (ECG). Note the large artifacts (E) on the intracardiac channel seen without ventricular activity (R, native QRS; V, pace output). These were caused by failing inner insulation in a bipolar coaxial lead system.

Oversensing need not result in pacing system inhibition. In the dual-chamber mode, if oversensing occurs on the atrial channel, it can result in ventricular pacing in response to the false P wave, which in a patient susceptible to retrograde conduction can precipitate an ELT as well. If the pacemaker was programmed to the triggered mode (which is one way to manage an oversensing problem that cannot be safely eliminated by programming to a less sensitive value), there would be periods of irregular pacing at a rate exceeding the programmed base rate in the presence of oversensing. Note that many of the biventricular devices use a trigger ventricular pacing function to pace on a sensed ventricular beat in an attempt to maintain resynchronization therapy.

Oversensing can be confirmed by use of the magnet or by reprogramming to an asynchronous mode. If the cause of the pauses or slow rate is oversensing, regular pacing will resume during asynchronous pacing. If a lead failure, open circuit, or other cause exists as the cause of lack of output, these maneuvers will not result in consistent asynchronous pacing. The use of EGMs and event marker telemetry can quickly identify the cause of the pause, whether an output pulse that failed to reach the heart or the presence of oversensing. In the case of oversensing, the telemetered EGM may provide a clue to the source of the extraneous signal. As demonstrated in Figure 29-30, this is useful for differentiating the causes of oversensing and deciding on the proper remedy. The normal refractory period timing cycle initiated by the oversensing may give the overt appearance of undersensing, because the native complex that coincides with the electrocardiographically invisible refractory period is not sensed (normal behavior), allowing for time-out of the pacing escape interval and delivery of an output pulse; this is termed functional undersensing (see Fig. 29-4). The output pulse, if delivered during a period of physiologic refractoriness initiated by the native cardiac depolarization, may not capture, a phenomenon appropriately termed functional noncapture (Fig. 29-32). The true cause of the observed behavior is readily appreciated with telemetered event markers, but it may be difficult to decipher from the freestanding ECG or rhythm strip.

Figure 29-32 ECG rhythm strip showing functional noncapture.

The problem in this case is the pacing system’s failure to sense the QRS complex. A subsequent pace output (V) occurs during the myocardial refractory period and should not be expected to be captured. The second-to-last QRS represents fusion (F) between the pace artifact and the native QRS.

Primary component failure of the pulse generator circuit is the least common of all the causes of pacing system failure, but it too can result in failure to output (Box 29-8). Patients who have been lost to follow-up may present with complete exhaustion of the power source (battery) and a completely inoperative and unresponsive system. As the power source begins to fail, slowing of the pacing rate and erratic behavior may occur (Fig. 29-33). This may also be associated with mode and rate changes with or without sensing problems.

Figure 29-33 ECG rhythm strip from pulse generator beyond elective replacement interval.

Not only is the pacing rate below the programmed rate of 70 bpm, but the device is not sensing properly. The leads were found to function normally after the device was replaced.

Several problems may be mistaken for pulse generator failure, resulting in a no-output situation. These are generally related to problems at implantation. If the lead is not positioned properly in the connector block of the pacemaker, or if an improper lead is used, the electrical connection may not be complete. This is still an open circuit, but in the presence of a technically normal pulse generator and normal lead. The two are simply not connected appropriately, which is a physician error at implantation. A transient situation has been reported in unipolar systems after wound closure. Air may be present in the pocket, especially if a large generator is replaced by a smaller one. Air is an effective insulator and prevents contact of the anterior anodal surface of the pacemaker’s case with the body’s tissues (this applies only to coated pacemakers). The result is an incomplete circuit until the air is absorbed. Subcutaneous emphysema has also been reported to be responsible for loss of anodal contact.^151,152 A similar pseudofailure occurs when a unipolar pulse generator is out of the pocket. If the stimulation impedance could be measured in each of these settings, it would be extremely high and well above the standard measurement for the particular lead being used. Event markers would report the release of an output pulse, even though the circuit was open, precluding effective pacing.

It is important to keep in mind that virtually all bipolar pacemakers may be programmed to function in the unipolar mode. Failure to output may also be seen with bipolar pacemakers that are connected to a unipolar pacing lead when programmed to pace in the bipolar polarity. This is more likely with IS-1 leads because the bipolar and unipolar connectors may appear identical. Most unipolar IS-1 leads have a protective metal band where the anodal screw of the pacemaker connector block is located to prevent inadvertent tightening of this screw from damaging the lead. This gives the appearance of a bipolar lead, even though no anodal conductor is present, and may be misleading (Fig. 29-34). Most polarity-programmable pulse generators allow the user to designate the lead type as “unipolar” in a separate field of the programming system. Then, when the standard programmed parameter page is accessed, the “bipolar” lead option is locked out, minimizing the chance that the device will be accidentally programmed to the bipolar configuration. Of course, this lockout feature only works if the lead type is programmed correctly. Some newer devices have the ability to verify that a bipolar circuit is present before allowing permanent programming to be completed to the bipolar polarity. This feature operates independently of any programmed lead type and also protects against programming to an ineffective polarity if the anode has failed because of lead fracture, malalignment of the lead in the connector, or failure to secure the anode set screw.

Figure 29-34 Unipolar and bipolar IS-1 connectors.

Top lead is an endocardial bipolar lead, and bottom lead is an epicardial unipolar lead. Note that both leads have a pin and ring at the connector end, leading to the possibility of confusion that the unipolar lead might actually be bipolar.

In dual-chamber systems, sensing an event on one channel triggers refractory periods on both channels. Oversensing on the ventricular channel initiates both a ventricular refractory period and a PVARP. Depending on its precise timing, the PVARP initiated by the oversensed ventricular event may result in failure to sense a native P wave—yet another example of functional undersensing. A similar phenomenon may be seen in single-chamber systems in the presence of relatively rapid intrinsic rhythms. Oversensing starts a refractory period that may coincide with a native complex. Although this same complex is otherwise perfectly capable of being sensed, the pacemaker’s refractory period precludes it being sensed, with the result apparent undersensing when the true cause of the problem is oversensing. This is most readily identified if the phenomenon occurs while the rhythm is being monitored with the simultaneous display of event markers.

Programming errors by the clinician also may result in no output or the appearance of no output. Programming to an ineffective pacing mode such as ODO will not allow any pacing to occur. A similar problem might be seen in a patient with complete AV block if programmed to the AAI mode. Likewise, programming the pace outputs to zero or extremely low values may cause the appearance of failure output.

Failure to Capture (Artifact Present)

With the advent of the newer ECG recording systems and the use of pulse artifact enhancement, the clinician must first be certain that the pace artifact seen is truly from the pacing system. The newer recording and monitoring systems have pacemaker pulse detectors that, when enabled, generate a discrete pacemaker pulse artifact in response to any high-frequency transient. Many extraneous signals can cause the enhanced ECG, telemetry system, or Holter recorder to place an erroneous pace artifact on the ECG. EMI (including programmer telemetry), loose ECG electrode connections, and subthreshold pulses, as seen with the minute ventilation sensors associated with normal pacemaker function, are examples of situations associated with spurious ECG artifacts (see Fig. 29-23). The use of event marker telemetry and EGMs can assist in differentiating the true pacemaker output from a recording system artifact, whereas retrospective analysis of a recorded rhythm may prove challenging.

A common cause of functional noncapture is undersensing, with a resultant delivery of the pacemaker impulse during the physiologic refractory period of the myocardium. Although the output does not elicit a depolarization, the problem in this case is related to sensing, not output, and should be evaluated as such (see Fig. 29-32). This is best identified as undersensing, but if one insists on commenting on capture, it is functional noncapture. No matter how high the energy content of the output pulse, the myocardium is physiologically refractory and incapable of being depolarized.

When true noncapture is documented, the patient’s safety must be ensured. If the patient is pacemaker dependent or is intolerant of the subsequent bradycardia associated with complete or intermittent loss of capture, the device should be programmed to its highest output. This can often be rapidly achieved by activating emergency VVI mode through the programmer. In most systems, even with a dual-chamber device, this programs the pacemaker to the highest available output in a unipolar output configuration. If consistent capture is still not secured, placement of a temporary pacemaker may be required. Almost all bipolar devices allow programming to the unipolar polarity. This may acutely correct the situation, particularly in the presence of an internal insulation failure, with a resultant short circuit or an open circuit associated with a fracture of the outer (anode) conductor. Although this maneuver eliminates the need for an emergency surgical intervention, it is a temporizing measure only, and a lead with a mechanical failure should be replaced.

Box 29-9 lists the causes of noncapture. The onset of noncapture in relation to implantation of the device and lead system can provide valuable clues to the cause. Loss of capture shortly after lead implantation suggests dislodgment, malposition, or perforation of the heart by the lead. Elevated thresholds that occur during the first several weeks after implantation were more common in the early days of pacing. Although the incidence of this problem has greatly decreased with the introduction of steroid-eluting electrodes and other electrode materials and designs, a significant rise in capture threshold may still occur. The capture threshold rise is attributed to the inflammatory reaction that results from two factors: (1) the pressure and thus trauma applied by the electrode to the endocardium and (2) a standard foreign body reaction. The peak of the early threshold rise usually occurs 2 to 6 weeks after implantation. Use of a higher output during this acute period minimizes the likelihood of noncapture until the threshold improves to its chronic level. Use of newer devices that can measure the capture threshold on a beat-to-beat or periodic basis can reduce the need for these higher outputs, yet provide the safety needed for the patient should the threshold rise. A rise in threshold more than 6 weeks after implantation is usually considered to be in the chronic phase of the lead maturation process. The threshold may rise steadily over time until noncapture occurs or until the threshold exceeds the pacemaker’s maximum output, typically referred to as exit block.^153–159 Older lead models and some epicardial leads are more likely to develop elevated capture threshold levels.^160–163 Some patients demonstrate repeated episodes of this phenomenon and have required multiple lead revisions or replacements. Children with epicardial implants are especially prone to this condition.^153,164 Steroid-eluting leads are useful in minimizing the development of exit block.^165–173 Anecdotal reports propose the use of high-dose systemic steroids to lower high thresholds, with dosing continued for months and then slowly tapered, but no prospective randomized trials have evaluated this pharmacologic intervention.^174–176

Causes of capture threshold rise occur in both the acute and the chronic phase. Any severe metabolic or electrolyte derangement can lead to acute threshold changes^177–188 (Box 29-10). These may occur in critically ill patients or during and immediately after cardiac arrest. Noncapture may occur secondary to the hypoxemia, acidemia, or hyperkalemia, rather than be the primary event leading to the arrest. In addition, some medications may affect capture thresholds, resulting in significant changes from the patient’s baseline^189–194 (Box 29-11). Beat-by-beat capture confirmation and periodic capture threshold measurement systems are now in widespread use. These have been reported to identify patients with transient or sustained late rises in capture threshold that would have resulted in loss of capture had the implanted pacemaker been programmed to a standard 2 : 1 safety margin.¹⁹⁵ As experience continues using systems with these capabilities, understanding of the factors that can affect chronic capture threshold will increase.

Although the clinician might expect that fracture of a lead would result in the absence of any visible pacing artifact, this is frequently not the case. The ECG may display a pace artifact with failure to capture, especially in unipolar systems. This occurs because the current passes across the gap in the conductor coil through the fluid in the lead; however, the resistance is high, and the delivered current is likely to be subthreshold. In unipolar systems, the anode (i.e., the pulse generator surface) is virtually always intact and provides enough of an electrical transient to trigger the artificial pace artifact circuitry of many monitoring systems. One can also see stimulus artifacts in the setting of a total disruption of both the conductor and the insulating sheath, which allows the output to be delivered to some local tissue, resulting in a stimulus on the ECG even though there is noncapture.¹⁴⁵

Although modern pulse generators attempt to maintain the programmed voltage output until the battery is exhausted, a reduction in the delivered voltage may occur before the elective replacement indicator is activated. At advanced stages of depletion of the battery, this may result in an ineffective pacing stimulus.

Latency is a finding that may be mistaken for failure to capture. Latency is defined as the delay between delivery of the pacing impulse and onset of the evoked potential or electrical systole.¹⁹⁶ Some antiarrhythmic agents, myocardial disease, or severe electrolyte disturbances (e.g., hyperkalemia) can cause latency (Fig. 29-35). A Wenckebach type of latency prolongation, leading to a noncapture and a repeat of the cycle, has also been reported.

Figure 29-35 ECG rhythm strip showing latency.

Note the prolonged period from the pace artifact until the evoked QRS complex. This patient had a serum potassium concentration of 7.1 mEq because of renal failure.

Failure to Sense

To appreciate undersensing, it is necessary to understand how the sense amplifier works and where in the cardiac cycle the sensing can and cannot occur. A prerequisite for proper sensing is a good-quality EGM; the signal must possess both an adequate amplitude and an adequate frequency content (slew rate) to be sensed properly. A signal of apparently adequate peak-to-peak amplitude may be greatly attenuated by the sense amplifier because of its poor slew rate. The resultant “filtered” signal may be of insufficient size to be recognized as a valid event.^197–202 Box 29-12 lists the causes of sensing failure.

It is often assumed that the pacemaker senses a cardiac event at the beginning or the peak of the P wave or QRS complex, as seen on the surface ECG. This is not the case. The surface P wave or QRS is a summation of all electrical events occurring at the cellular level over a period of time (normally 80 to 120 msec, but may be longer due to conduction system disturbances). The intracardiac EGM as recorded from a pacing electrode appears as a large-amplitude, relatively rapid deflection (see Fig. 29-10). If one did not know that the recording was an EGM, an atrial signal could be easily mistaken for a QRS complex presumed to have been recorded with a standard surface ECG. It is therefore helpful to record both the atrial and the ventricular EGM simultaneously, the EGM with the simultaneous surface ECG, or telemetered event markers (nearly all pacing systems now provide all these simultaneously). This event, the intrinsic deflection of the QRS or P wave, is generated as the wave of depolarization passes by the electrode. This is the complex that is sensed, even if it occurs well after the onset of the native depolarization. An excellent example of this delayed sensing is in the patient who has right bundle branch block and a pacing electrode in the right ventricle. The native depolarization occurs late in the right ventricle because of the conduction delay. The pacemaker does not sense this event until that depolarization reaches the pacing lead, which is very late in the timing of the surface QRS complex. If it is assumed that sensing occurred earlier, this may be incorrectly labeled “late sensing.” Measurements made from the onset of the QRS complex can misinterpret the timing cycle by more than 100 msec. Therefore, an ECG with a pacing artifact in the QRS (pseudofusion and fusion) can occur in the presence of absolutely normal sensing^203,204 (Fig. 29-36).

Figure 29-36 ECG rhythm strip showing pseudofusion beats.

Because the premature ventricular contractions arise from the left ventricle (right bundle branch block morphology), right ventricular depolarization occurs late within the overall cardiac activation sequence. The QRS complexes cannot be sensed by the pacemaker until late into the surface QRS, resulting in pseudofusion complexes when the escape interval “times out” before the intrinsic deflection has occurred. In these cases, even though the ventricle had started to depolarize, the pacemaker was not yet capable of sensing it.

The refractory and blanking periods are integral to understanding sensing and whether a native complex is even potentially capable of being sensed. A blanking period is a time during which no sensing at all will occur on a given channel. A refractory period is similar to a blanking period, but events occurring during a refractory period may be used for purposes other than resetting escape intervals. These “sensed but not used” events can be used by the device to detect noise on a lead, or detection of atrial events for initiation of mode switching. Long programmed refractory and blanking periods can cause undersensing of events occurring within these periods (Fig. 29-37). This might occur in the case of a closely coupled PVC or PAC or very rapid intrinsic rhythms. Although there may be resultant competition, this is not a lead or pulse generator malfunction, in that both are functioning properly and in accord with their design. This is clinically undesirable and reflects inappropriate programmed settings in the pacemaker, a behavior amenable to correction by programming. This is also termed functional undersensing, because the pacemaker is not capable of sensing at the time the native signal occurred. On the other hand, the failure to sense the native complex allows the timing period to complete, with the resultant release of an output pulse that may prove to be ineffective because it was delivered at a time of physiologic refractoriness. Functional undersensing usually results in functional noncapture. A classic example of this phenomenon was the committed DVI pacing system, an early-generation (and now obsolete) AV sequential pacing system that was not capable of sensing on the atrial channel but, with release of an atrial output, was committed to release a ventricular output even in the presence of a native R wave.^203,222

Figure 29-37 ECG rhythm strip showing undersensing of QRS complexes in patient with atrial fibrillation.

The device is programmed to VVI with a demand rate of 70 bpm and a refractory period of 475 msec. Because of a long programmed ventricular refractory period, this does not represent a pacemaker malfunction. The pacemaker will not respond to intrinsic events during the refractory period, a phenomenon termed functional undersensing. The arrows indicate paced beats.

As in noncapture, the onset of undersensing in relation to implantation of the pacemaker and lead system may direct the clinician to the correct diagnosis of the malfunction. Undersensing that occurs close in time to implantation should lead the physician to suspect dislodgment, malposition, or perforation of the electrode. Problems occurring in chronic systems are frequently caused by mechanical problems with the lead or, if the lead is normal, by programming errors. Occasionally, problems are attributable to a change in the morphology of the intrinsic cardiac signal associated with myocardial disease progression. Inappropriate programmed settings do occur, particularly when sensitivity threshold testing has been performed in a nontemporary manner during follow-up. The clinician may forget to return the sensitivity to the appropriate setting and leave it at or above the threshold setting. Occasionally, the sensing threshold may show a slow decline in the amplitude of the native signal as the lead-tissue interface changes over time.^205–212

The vector of depolarization can affect the amplitude and slew rate of the EGM. Any change in the vector may result in significant changes in the sensing threshold. If a patient with a ventricular pacemaker has a myocardial infarction, develops a bundle branch block, or has PVCs, the intracardiac signal may be insufficient to allow sensing at a previously appropriate setting^206–208 (Fig. 29-38). The same may be seen with atrial implants and PACs. Another cause of vector change is movement of the lead’s position associated with respiration. Marked changes in the intracardiac signal can be documented in some patients as the angle of the lead relative to the heart changes with body position or diaphragmatic motion^209–212 (Fig. 29-39). Evaluation of the quality of the intracardiac signal associated with sensing failure is greatly facilitated by access to the telemetered EGM.

Figure 29-38 Intracardiac electrogram (ICEGM) and surface ECG (I, II, III) of intrinsic beats and PVC.

In this example, the premature ventricular contraction (PVC) electrogram has a peak-to-peak amplitude that is significantly lower than the normal beats and may lead to undersensing.

Figure 29-39 Intra-atrial electrogram and surface ECG.

Note the profound variation of the intracardiac signal without apparent change in the surface P-wave morphology. The variation was synchronous with respiratory movements. This situation may be seen in either the atrium or the ventricle. ICEGM, Intracardiac electrogram.

As the pacemaker’s battery begins to become depleted, erratic behavior may occur, including persistent or intermittent undersensing. Failure of a circuit component may also cause sensing failure. Rarely, the reed switch that activates the magnet mode may stick in the closed position, causing (in most cases) asynchronous pacing. For the same reason, when reviewing a rhythm strip, the clinician must know whether the magnet was over the device when the recording was made. Asynchronous pacing may also be seen during the interference or noise mode function of the device. The latter obligates pacing when high-frequency, nonphysiologic electrical signals are sensed by the pulse generator (Fig. 29-40). Typically, this is defined as multiple sensed events occurring within the terminal portion of the ventricular refractory period. The device then paces asynchronously to protect the pacemaker-dependent patient against inappropriate device inhibition.^142,213

Figure 29-40 Noise response pacing.

Tracing was recorded from a patient with a St. Jude Medical pacemaker during exposure to electromagnetic interference. Note that the device goes from a normal tracking behavior to an asynchronous pacing behavior during the EMI. It then returns to normal function once the interference terminates. This feature is designed to prevent asystole in a pacemaker-dependent patient.

One frequently misinterpreted reason for failure to sense involves pacemakers that use impedance plethysmography as a rate modulation sensor. This applies primarily to minute ventilation sensors, but now also to “lung wetness” monitoring systems. The system uses the pacemaker’s case as the anode for the impedance-sensing impulses. Therefore, when the case is out of the pocket, much electrical noise is generated because the system is not grounded. This may result in asynchronous pacing until the device is placed into the pocket. If the clinician is not aware of this idiosyncrasy, a malfunction of the pulse generator might be inappropriately diagnosed.

The use of external or internal defibrillation can result in temporary or permanent loss of the sensing function. This may result from transient saturation of the sensing amplifier pacemaker, circuit damage, or lead-myocardium interface damage. This phenomenon occurs when the large electric current delivered by the defibrillator is diverted from the pacemaker’s circuitry to the pacing lead by way of Zener diodes, which are included to protect the pulse generator from large voltage surges. The result may cauterize the myocardium and thereby reduce or eliminate the local EGM. It may also cause a marked elevation in the capture threshold or result in exit block. Capacitive coupling is another mechanism that may allow a current to be induced into the pacing lead, by a defibrillation shock or nearby electrocautery use.^214–216 Exposure to magnetic resonance imaging (MRI) may result in inductive coupling with the lead system, induce a current, and cause the same problem.

Crosstalk Inhibition

Crosstalk is a potentially catastrophic form of ventricular oversensing that can occur in any dual-chamber mode in which the atrium is paced and the ventricle is sensed and paced. Consider that the common output voltages for the atrial channel range from 2.5 to 5 V, or 2500 to 5000 mV. The ventricular lead is just several centimeters away from this atrial lead and is “looking” for an electrical event of about 2 mV, which would represent a sensed ventricular depolarization. If the ventricular channel senses the pacing impulse delivered to the atrium, it interprets the event as an R wave. The ventricular output is inhibited, and the patient is left without pacing support. If crosstalk occurs in a patient with complete AV block, ventricular asystole could result. In the absence of an escape rhythm, only paced P waves would be visible.^217–220 Conversely, crosstalk is more difficult to detect when AV nodal conduction is intact. In a system that uses ventricular-based timing, the result may be an acceleration of the paced atrial rate with ventricular output inhibition, even in the presence of a first-degree AV block that exceeds the programmed AV delay.²²⁰

As demonstrated in Figure 29-41, crosstalk can be recognized by an acceleration of the atrial paced rate in a non-sensor-driven dual-chamber system using ventricular-based timing. Note that ventricular sensing occurs virtually simultaneously with atrial output. The AV delay is immediately terminated with ventricular sensing, and the atrial escape interval (AEI) is started. Therefore, the pacing interval with this type of timing is equal to the AEI (programmed pacing interval minus AV delay) plus the ventricular blanking period. Rate-modulated pacing systems that use ventricular-based timing have a rate acceleration that is seen at the base pacing rate. Crosstalk may not be readily diagnosed in patients with rate-modulated pacing systems and intact AV nodal conduction if the increased atrial pacing rate is believed to be caused by the sensor.

Figure 29-41 Crosstalk.

A, Electrocardiogram (ECG) rhythm strip showing crosstalk in a patient with complete atrioventricular (AV) block, with a ventricular-based pacing system. Note the presence of an atrial output without a ventricular output. A comparison of the atrial escape interval (AEI) from the first until the second atrial pulse with that from the second until the third atrial pulse demonstrates that the pacing interval is shortened by the A-V interval minus the blanking period (120 msec). B, ECG strip from the same patient with event markers during crosstalk. Note that the ventricular sensing (R) occurs immediately after the atrial output pulse (A). Safety pacing was turned off. C, ECG rhythm strip showing crosstalk in a patient with intact AV conduction. Because this pacemaker operates with ventricular-based timing, the AEI resets with each atrial output, elevating the frequency of atrial stimulation. D, ECG rhythm strip and event markers during crosstalk and ventricular safety pacing enabled. Note that, despite the programmed setting of the AV interval to 200 msec, the event markers demonstrate the ventricular stimulus at 125 msec after the atrial output pulses. IEGM, Intracardiac electrogram.

Pacing systems that use atrial-based timing do not exhibit this rate increase. Although ventricular sensing of the atrial output inhibits the ventricular output, the portion of the AV interval not used is added to the AEI, maintaining atrial pacing at the programmed interval (Fig. 29-42). Without access to event marker telemetry, ventricular output inhibition from crosstalk cannot be differentiated from failure to output, as with an open circuit.

Figure 29-42 ECG rhythm strip showing crosstalk with atrial-based timing in pacemaker.

Note that the atrial pacing rate remains at the programmed rate of 70 bpm. The atrioventricular interval was programmed to 175 msec.

Several factors increase the likelihood of crosstalk (Box 29-13). Programmed settings that make the ventricular channel more sensitive or that increase the atrial energy output predispose to this problem. Insulation failure on the atrial lead also increases the chances of crosstalk by increasing the pulse charge and energy. Certain pacemaker models are more prone to this phenomenon than others, related to differences in circuit designs and the degree of isolation between the two channels within the pulse generator.²¹⁷ Crosstalk may occur internally, within the circuitry of the pulse generator, as a result of external oversensing between the pacing leads. Unipolar systems are also significantly more prone to crosstalk because of the larger electrical signal, as sensed by the ventricular channel.^218–223

Prevention of Crosstalk

Prevention of crosstalk is approached by proper programming of the device and proper lead placement. Atrial amplitude and pulse width settings that provide an appropriate safety margin without being excessive and ventricular sensitivity settings that are not unnecessarily sensitive are helpful. Newer circuit designs are much more resistant to this problem. Even under ideal circumstances, however, crosstalk may still occur.

The use of an additional ventricular refractory period, known as the ventricular blanking period (VBP), is fundamental to prevention.²²¹ The VBP begins with the atrial output and usually lasts from 12 to 120 msec (the “ultimate blanking period” was the now obsolete DVI-committed mode, during which no sensing occurred on the ventricular channel at any time during the A-V interval after an atrial output pulse). In most devices, this is a programmable parameter. During the VBP, nothing (including the atrial output) can be sensed on the ventricular channel. Short VBPs help to prevent the undersensing of intrinsic ventricular activity but may allow crosstalk to occur. Long VBPs virtually ensure the prevention of crosstalk but may cause intrinsic events to not be sensed, a potential problem if the native beat is a late-cycle PVC and the A-V interval is programmed to a long length. Functional failure to sense a PVC, because the intrinsic deflection of the PVC coincided with the VBP, may be followed by a ventricular output beyond the myocardial refractory period and on the vulnerable zone of the T wave. This might result in the induction of a ventricular arrhythmia, which could also occur with atrial undersensing if a normally conducted QRS complex falls into the VBP. The latter event would not be detected, and an unnecessary stimulus would be delivered (Fig. 29-43).

Figure 29-43 ECG rhythm strip showing undersensing of native QRS complex.

The device was programmed to the DVI mode, so there is no atrial sensing. When the atrial escape interval expires, the atrial output coincides with the onset of the native QRS complex. The ventricular blanking period is triggered by the release of the atrial output; this precludes ventricular sensing for 38 msec after the atrial output. If the intrinsic deflection of the native QRS coincides with the ventricular blanking period, it will not be sensed, and a ventricular output will be delivered at the end of the programmed atrioventricular interval.

Prevention of Crosstalk Sequelae

The earliest approach to prevention of ventricular output loss as a result of crosstalk was the introduction of the DVI-committed mode (DVI-C). In this version of DVI, ventricular sensing is completely disabled during the AV interval that follows an atrial output pulse. After termination of the A-V interval, a ventricular output pulse is committed whether or not an intrinsic ventricular event has occurred. Thus, with DVI-C, there can never be an inhibition of the ventricular output by an atrial output. Although this approach worked, it resulted in confusing ECGs and wasted energy because of the delivery of unnecessary pacing impulses.^203,222 It is also not practical for DDD devices for similar reasons. With the development of the DDD mode, however, manufacturers were able to provide brief refractory periods (blanking periods) starting with the atrial output. Even blanking periods, however, are not a guaranteed prevention for ventricular output inhibition associated with crosstalk.

To ensure patient safety, a unique circuit/algorithm has been integrated into the ventricular channel. The most common approach is the use of a safety output pulse when crosstalk is suspected by the device. This is referred to by multiple names, depending on the manufacturer of the device (e.g., safety pacing, nonphysiologic AV delay, ventricular safety standby). This method uses a brief sensing period (crosstalk sensing window) after the blanking period. Any electrical event sensed during this crosstalk sensing period is assumed to be crosstalk. A ventricular output is then triggered to be delivered after a shortened AV delay^223,224 (Fig. 29-44). Other events can cause a safety output pulse to occur, including premature ventricular beats, premature junctional beats, and normally conducted ventricular events that occur after undersensed intrinsic atrial beats (see Fig. 29-43). After one of the latter events, if the safety output pulse were to be delivered using a long A-V interval, there would be a potential for delivering a tightly coupled and potentially arrhythmogenic stimulus to the ventricle. Therefore, a shortened A-V interval is used (typically 100-120 msec) to deliver the stimulus harmlessly into the depolarizing ventricle while maintaining the ability to rescue the patient in the presence of actual crosstalk. In the presence of a long programmed AV delay and intact AV nodal conduction, crosstalk is signaled by AV pacing at the abbreviated AV delay. Reducing the atrial output, reducing the ventricular sensitivity, or lengthening the VBP should reduce, if not totally eliminate, episodes of crosstalk.

Figure 29-44 ECG rhythm strip showing safety pacing.

Sensing is occurring on the ventricular channel during the crosstalk sensing window. This causes a ventricular output to be delivered at a shortened atrioventricular (A-V) interval. In a ventricular-based timing system, the pacing interval is shortened by the difference between the programmed A-V interval and the safety-paced A-V interval, because the atrial escape interval is reset by any sensed or paced ventricular event. This results in an AV paced rate of 93 bpm, even though the device is programmed DDD at 75 bpm and is not capable of rate modulation.

As in crosstalk, delivery of the safety output pulse causes a shortening of the base pacing interval because of the shortened A-V interval. If the baseline A-V interval is programmed to the same duration as that of the safety output pulse (this has become very common in biventricular devices when treating patients with congestive heart failure), or if a rate-modulated or differential A-V interval is present, the presence of safety pacing (i.e., crosstalk) may be difficult or impossible to discern on the standard ECG. The use of event marker telemetry documents the presence of crosstalk or safety output pulses in these situations (see Fig. 29-44).

Delivery of a safety output pulse does not prevent crosstalk. It is designed to prevent the sequelae of crosstalk by preventing ventricular asystole. It is unwise to allow a patient to use this backup mechanism continually. If crosstalk is suspected or documented, the clinician should pursue the potential causes and make the necessary adjustments to terminate this condition. In some situations, such as failure of the lead’s insulation, programming changes may not be adequate, and surgical intervention may be required.

Pacemaker-Mediated Tachycardia

Any undesired rapid pacing rate caused by the pulse generator or by an interaction of the pacing system with the patient may be considered pacemaker-mediated tachycardia (PMT). This has classically been associated with the endless-loop tachycardia (ELT) seen in dual-chamber devices. PMT, however, should be assumed to be a more general term. The following are several conditions that are well-documented causes of PMT.

Runaway Pacemaker

The runaway pacemaker is a pacemaker malfunction that may occur in single-chamber or dual-chamber pacing systems. It is usually the result of at least two separate component failures within the pulse generator. The result is the rapid delivery of pacing stimuli to the heart, with the potential for inducing lethal arrhythmias, such as ventricular tachycardia or fibrillation. All modern devices incorporate a runaway protect circuit that prevents stimulation above a preset rate, typically between 180 and 200 bpm. Although extremely rare with modern pacing devices, this represents a medical emergency (Fig. 29-45). Emergent surgical intervention to replace the device or, if all else fails, cut the lead, must be performed. Obviously, a patient who is pacemaker dependent presents a new challenge as soon as the defective pacing system is disabled.^225,226

Figure 29-45 Example of “runaway pacemaker.”

This device was subjected to direct, unshielded therapeutic irradiation during treatment of a breast carcinoma. The patient presented to the clinic for routine follow-up and had a normal presenting rhythm strip, as shown. She had been programmed to DDDR from 70 to 120 bpm. On placement of the magnet over the device, pacing began at the runaway protect rate of 180 bpm. The patient was taken to the operating room, and the device was replaced emergently.

Sensor-Driven Tachycardia

Sensor-driven tachycardia is a rapid heart rate occurring in rate-modulated pacemakers. It can occur with any type of sensor-driven pacing system (Box 29-14). Interaction between the patient and external stimuli can cause the rate modulation system to overreact and pace at a high rate. The most common cause of this PMT type is inappropriate programming of the sensor parameters. In piezoelectric devices, a threshold setting that is too low or a slope setting that is too high results in high pacing rates for low levels of activity. It also exposes the patient to increased pacing rates for nonphysiologic events, such as riding in a car, exposure to loud noise or music, or sleeping on the ipsilateral side to the device implant.^227–237

Box 29-14
Causes of Sensor-Driven Pacemaker Tachycardia

Vibration sensor: piezocrystal

Pressure on device

• Patient lying on device (as during sleep)

• Tapping on device

Submuscular implantation

Loud, low-frequency music

Bumpy ride, as in a tractor, lawn mower, or helicopter

Use of vibration-generating tools

Impedance-based sensors—minute ventilation

Hyperventilation

Electrocautery

Environmental 50- to 60-Hz electrical interference

Evoked QT interval

Rate-dependent QT-shortening spiral

Sympathomimetic drugs

One unpublished incident concerned a patient with a bipolar piezoelectric VVIR device. The emergency medical squad had been summoned to treat an unconscious patient. They arrived to find the patient with seizure activity. On viewing the ECG, they found the patient’s condition to be wide-complex tachycardia at 150 bpm. The patient received repeated direct-current cardioversion for this tachycardia, without reversion to sinus rhythm. In this patient the seizures were caused by epilepsy, and the vibration-based sensor responded to the seizure by pacing at its upper rate. No spikes had been noted on the ECG because the output configuration was bipolar, which resulted in a diminutive artifact on the monitor (Fig. 29-46).

Figure 29-46 ECG rhythm strip of pseudo–ventricular tachycardia.

Caused by rapid ventricular pacing using a bipolar device. Note that the pace artifact is not well seen, leading to this potential misdiagnosis.

In the past, pacing systems responsive to central venous temperature changes might have responded to a fever spike with upper rate pacing. This necessitated design of a “nulling” feature, which restores a more appropriate pacing rate after a specified time at the upper rate has occurred. Devices that respond to changes in the evoked QT interval also use this nulling feature, although for different reasons. As the QT interval shortens because of catecholamine increases and changes in the autonomic nervous system, the pacing rate is increased by the device; however, this may cause further shortening of the QT interval by a rate-dependent mechanism. The latter change further increases the pacing rate, which may then spiral to the upper rate limit.

Devices that use thoracic impedance plethysmography to determine changes in minute ventilation may pace at the upper rate during marked hyperventilation, use of electrocautery, hyperventilation during anesthesia induction, and when close to an electrical power supply. Therefore, it is strongly recommended that this type of rate-modulating feature be disabled before the patient undergoes any surgical procedure (even if electrocautery is not used, because much electrical equipment is present in most operating room suites), treatment with a mechanical ventilator, or treatment in a critical care unit.²³⁶

Myopotential Tracking

Myopotential tracking is caused by oversensing of muscle potentials by the atrial channel in a dual-chamber pacing system capable of P-wave tracking. This has become much less common with the use of bipolar sensing, which is preferred by most implanting physicians. It is a greater problem with unipolar dual-chamber systems. A unique sensing modality—“combipolar” sensing, which is sensing between unipolar atrial and ventricular electrodes for a form of wide-dipole bipolar sensing with the entire system restricted to the heart—has the potential to reduce the incidence of atrial myopotential oversensing.²⁴⁵ Myopotential tracking occurs when the atrial channel senses the electrical activity of the muscle underlying the pulse generator. Those same signals, if sensed on the ventricular channel, would result in myopotential inhibition. The atrial sensitivity setting is usually more sensitive than in the ventricular channel, and myopotential sensing is more likely in the unipolar configuration. When myopotentials are sensed on the atrial channel, the atrial output is inhibited, and a ventricular output is triggered at the end of the AV interval. This pseudo–P wave repeatedly triggers the ventricular output, and ventricular pacing may occur up to the programmed maximum P-wave tracking rate (Fig. 29-47). If the myopotentials are of sufficient amplitude, they may be sensed on the ventricular channel, in which pacing system inhibition may also occur.^88,89,120

Figure 29-47 ECG rhythm strip of myopotential tracking.

The electrical activity of the muscle adjacent to the pacemaker case in a unipolar system is sensed by the atrial channel. This causes repeated initiation of the atrioventricular interval up to the maximum tracking rate. Higher-amplitude myopotentials may also cause inhibition of the ventricular channel.

Endless-loop Tachycardia

The classic form of PMT is ELT, which can occur in dual-chamber pacemakers that are capable of atrial tracking modes, most often DDD or VDD. Only patients who are capable of retrograde ventriculoatrial (VA) conduction through the AV node or an AV accessory pathway are capable of sustaining this rhythm. The mechanism is identical to any macro-reentrant tachycardia in which two electrical pathways exist between the atria and ventricles. Pacemaker ELT is classically initiated by a PVC. The depolarization is conducted retrograde to the atria. If the PVARP has ended and the retrograde complex is of sufficient amplitude to be sensed, the atrial channel senses the event and initiates an A-V interval. At the end of the A-V interval, the pacemaker delivers a stimulus to the ventricle, and the loop is reinitiated (Fig. 29-48). Although PVC is the classic cause of PMT initiation, any situation that results in AV dissociation, allowing a ventricular depolarization to occur without a normally coupled atrial-paced or atrial-sensed event, may begin the loop (Box 29-15). Examples include atrial noncapture, oversensing, and undersensing. In addition, if the programmed AV interval is long (some units allow programming of an interval of up to 400 msec), it may be possible for the AV node to recover in time to conduct the subsequent ventricular-paced event in a retrograde direction, thus initiating another ELT (Fig. 29-49). The absence of anterograde AV nodal conduction does not rule out retrograde conduction over the AV node or a concealed AV accessory pathway. Patients may also exhibit intermittent retrograde conduction or have variations in the retrograde conduction time based on their sympathetic tone and catecholamine status.^246–255

Figure 29-48 Classic initiation of endless-loop tachycardia by PVC in patient with retrograde AV conduction.

A, Premature ventricular contraction (PVC) occurs. B, The event is conducted retrograde to the atrium, as seen by the inverted (retrograde) P wave (RP). C, The retrograde event is sensed by the atrial channel, which causes an atrioventricular (A-V) interval (AVI) to be started. D, At the end of the AVI, a stimulus is delivered to the ventricle (if the maximum rate limit is not violated), resulting in retrograde conduction and the perpetuation of this cycle. AVN, Atrioventricular node; LA, left atrium; LV, left ventricle; PVARP, postventricular atrial refractory period; RA, right atrium; RV, right ventricle.

Box 29-15
Endless- Loop Tachycardia (ELT) Summary

Initiation

Premature ventricular ectopic beat

Atrial undersensing

Atrial oversensing

Atrial noncapture

Long programmed AV delay with echo beats

Prevention

PVARP

PVARP extension after ventricular ectopic beat detection

Differential atrial sensing

Rate-responsive AV delay allowing longer base-rate PVARP

Adaptive PVARP

High maximum tracking rate (MTR): retrograde fatigue

Atrial pace on PVE detection

Detection

P tracking at upper rate limit

P tracking at intermediate rate

Atrioventricular (A-V) interval modulation

Discrepancy between sensed atrial rate and sensor-indicated rate

Termination

PVARP extension

Withhold of ventricular output

Withhold of ventricular output with early atrial pace

Atrial pace during ELT

AV, Atrioventricular; PVARP, postventricular atrial blanking period; PVE, premature ventricular event.

Figure 29-49 Initiating events leading to retrograde P-wave conduction and potential for endless-loop tachycardia.

A, Atrial noncapture causes ventricular stimulation without the atrioventricular (AV) node being blocked by an anterograde event. B, Atrial oversensing inhibits atrial output and provides just a ventricular output, with the same results as in A. C, Atrial undersensing results in a long A-V interval, allowing the AV node time to recover and conduct retrograde. D, A long programmed A-V interval may also allow the AV node to recover.

The rate of ELT depends on the conduction velocity and refractory period of the retrograde AV pathway. If the retrograde AV conduction is the same as or shorter than the upper rate interval (but not shorter than the PVARP), the tachycardia rate is at the programmed upper rate. If the retrograde AV conduction time is slower than the upper rate interval, the tachycardia rate is below the upper programmed rate (Fig. 29-50). Therefore, although the rate of the ELT can never exceed the programmed upper rate of the pacemaker, it may be lower.^254,256 When the rate is lower than the MTR, the rhythm has been termed a balanced ELT.

Figure 29-50 Endless-loop tachycardia (ELT).

A, Electrocardiogram (ECG) rhythm strip of ELT with fast ventriculoatrial (VA) conduction, which resulted in a rapid ELT at the upper programmed rate of 155 beats per minute (bpm). The lower rate was 60 pulses per minute (ppm); the sensed atrioventricular (A-V) interval, 150 msec; the postventricular atrial refractory period (PVARP), 275 msec; and the VA time, 300 msec. B, ECG rhythm strip of ELT with slow VA conduction, which resulted in ELT below the upper programmed rate. Because of the slow retrograde conduction, the ELT rate is significantly slower than the programmed upper rate. The lower rate was 60 bpm; the upper rate, 145 bpm; the AV interval, 165 msec; and the PVARP, 250 msec. P, Retrograde P wave; V, ventricular output.

Prevention

The main defense against ELT is the use of an appropriate PVARP interval. During the PVARP, the atrial channel cannot respond to the retrograde depolarization that would initiate the ELT. This is independent of the source of that atrial event. If retrograde conduction is present during implantation or follow-up, it is a simple matter to measure the VA time and program the PVARP to an interval that is longer.²⁵⁰ Many patients, however, exhibit only intermittent VA conduction, making an accurate assessment difficult. The major limitation of using a long PVARP is that it limits the upper tracking rate of the device (2 : 1 blocking rate). Some patients in our clinics have VA times in excess of 430 msec. If a PVARP of 450 msec and an A-V interval of 150 msec are used, the total atrial refractory period (TARP) is 600 msec. This causes 2 : 1 blocking at an atrial rate of 100 bpm, which is much too low for most patients.

One solution to this problem is the use of ELT prevention algorithms.^257,258 The most common algorithm is PVARP extension on PVE detection. A premature ventricular event is defined as a ventricular sensed event that is not preceded by an atrial paced or atrial sensed event. When a PVE is detected, the device prolongs PVARP by a fixed or programmable value for the next cycle. It allows a shorter baseline PVARP to be used, with associated higher 2 : 1 blocking rates. A variation on this technique is to initiate one DVI cycle after the PVE. This is the ultimate PVARP extension, because the atrial channel remains refractory to sensing throughout the entire AEI. A problem with most of the PVARP extension algorithms is that functional atrial noncapture may occur. Because there was an atrial output, the system returns to the shorter PVARP, thus allowing for retrograde conduction to be sensed after the AV paced cycle, simply postponing the ELT for one cycle. In the latter case, the pacemaker does not maintain the long PVARP because it has delivered an atrial stimulus (although ineffective) before a ventricular event. The subsequent ventricular paced event then leads to ELT (Fig. 29-51). A refinement to the standard PVARP extension algorithm is to extend the atrial alert interval at the end of the extended PVARP.²⁴⁶ On release of the atrial output at the end of the longer AEI, the atrial myocardium will have recovered, allowing for capture, thus precluding retrograde conduction on the next ventricular cycle.

Figure 29-51 Dual-chamber pulse generator programmed with PVARP extension on premature ventricular event detection.

Top tracing shows behavior of pacemaker with magnet application. The atrioventricular (AV) delay is programmed to a long value to allow functional single-chamber atrial pacing. Bottom tracing shows demand mode. A ventricular premature complex/contraction (PVC) occurred (arrow), causing postventricular atrial refractory period (PVARP) extension. The retrograde P wave coincided with the extended PVARP and, appropriately, was not sensed and not tracked. When the atrial output was delivered at the end of the atrial escape interval, it was ineffective (functional noncapture), because the native atrial depolarization had rendered the atrial myocardium physiologically refractory. The delivery of the atrial output caused the pacemaker to shorten the PVARP with the next ventricular paced event, by which time atrial refractoriness had resolved; retrograde conduction followed the paced ventricular complex and initiated an endless-loop tachycardia (ELT). Thus, the PVARP extension simply postponed the initiation of the ELT rather than preventing it. Data from Cosmos Model 283-01 (Sulzer Intermedics, Angleton, Texas, now Boston Scientific).

(From Levine PA, Selznick L: Prospective management of the patient with retrograde ventriculoatrial conduction: prevention and management of pacemaker mediated endless loop tachycardias, Sylmar, Calif, 1990, Pacesetter Systems.)

Newer approaches have been made possible by the advent of sensor-driven pacing systems. With a DDDR system, even though the MTR may be limited by a long PVARP, faster AV sequentially paced rates may be possible through the sensor. Even though atrial sensing does not occur during the PVARP, sensor-driven atrial pacing may occur. Another alternative in some DDDR pulse generators is to use an adaptive PVARP that is long when the sensor determines that the patient is at rest. When the patient’s sensor-indicated rate increases along with the need for higher heart rates, the PVARP is shortened proportionately, allowing higher rates before a 2 : 1 block upper rate behavior occurs. This approach allows the device to discriminate between early atrial depolarizations at rest (possibly retrograde beats or PACs) and sinus tachycardia with exertion.^259–261 Even in the absence of rate modulation, rate-responsive AV delay algorithms allow a shorter TARP and therefore a higher MTR, even with a fixed, long PVARP.

A comparison of P-wave morphology can allow discrimination between sinus and retrograde atrial depolarizations in some cases; this is referred to as differential atrial sensing.^262–266 Mapping of the atrium may allow positioning of the lead at a site that has a large P wave when generated by the sinus node and retrograde P waves that appear small in comparison. The pacemaker sensitivity may be programmed to allow sensing of the anterograde P wave but not the retrograde P wave. Therefore, ELT cannot be initiated or sustained because the loop cannot be completed. Although the retrograde P wave is usually of lower amplitude than the anterograde P wave, this is not always the case; the individual situation needs to be evaluated before any attempt to use atrial sensitivity as a discriminating factor. In addition, the atrial EGM may decrease in amplitude as the sinus rate increases under physiologic stress. Programming a reduced atrial sensitivity to facilitate discrimination between anterograde and retrograde P waves could compromise sensing of rapid intrinsic atrial rhythms.²⁴⁶

Another preventive mechanism has been to program a high MTR, in the hope of inducing retrograde fatigue and spontaneous termination of any ELT that is allowed to start. Careful assessment of the patient’s ability to conduct retrograde should be performed, however, because some patients have superior retrograde conduction (compared with anterograde conduction) and are able to sustain extremely rapid ELTs.²⁴⁶

Detection and Termination

The most difficult aspect of applying an ELT termination protocol is having the device determine whether ELT is present as opposed to a native atrial tachycardia that is being appropriately tracked. This has been approached differently by different manufacturers. Some criteria used are simple and others more complex. The following are some of the most common techniques for identifying the presence of ELT.

The first criterion is sustained P-wave tracking at the upper rate. This is the most common method of ELT detection. If the pacemaker tracks the P waves at the upper rate for a specific number of intervals, then the PVARP is prolonged for one cycle, DVI is used for one cycle, or the ventricular output is withheld. The number of events at the upper rate that trigger an ELT termination attempt depends on the particular pulse generator. Some algorithms use a fixed number of beats; in other systems, this is a programmable option. The method of ELT termination may also be programmable (PVARP extension, DVI cycle, or atrial pace).^267,268

The second method used to identify ELT is sustained P-wave tracking at a specific rate that is lower than the upper rate. As previously noted, not all ELTs occur at the upper programmed tracking rate. An ELT rate is always limited by the slowest point in the loop. If the patient has a VA conduction time that is longer than the upper tracking interval, the resulting ELT is slower than the programmed upper tracking rate. This manifestation of ELT is not identified by the sustained upper rate method. Some devices allow initiation of the ELT termination algorithm at a P-wave tracking rate that is lower than the upper tracking rate. This lower rate is typically a programmable option, to provide for termination of ELT in this subset of patients.

The third approach to identifying ELT is modulation of the A-V interval during sustained P-wave tracking at or above a specific rate. The technique is based on the observation that the VA conduction time in an individual patient is relatively constant. When the device is P-wave tracking at a high rate, the system first assesses the stability of the retrograde interval. The A-V interval is then either shortened or lengthened, and the effect on the subsequent retrograde interval is assessed. If an ELT is present and the V-pace to P-sense interval remains constant (i.e., atrial cycle length is shortened by shortening of A-V interval), the device confirms ELT and initiates a termination algorithm (Fig. 29-52), either by withholding the ventricular output or by delivering an atrial output after a period sufficient to allow for atrial recovery.^260,261 However, if the V-pace to A-sense interval changes by more than a preset amount (i.e., atrial cycle length is unaffected by A-V interval shortening), ELT is not confirmed, and the device continues to track the atrium normally.

Figure 29-52 Diagram of endless-loop tachycardia (ELT) detection scheme.

A, Rapid tracking of sinus tachycardia. The device causes the atrioventricular interval (AVI) to be varied. When the interval is shortened (AVI modified), there is a prolongation of the V-P interval (V-P2 is longer than V-P1) because the sinus node rate is not affected. B, ELT caused by tracking of retrograde P waves. The device again causes the AVI to be varied. In this case, shortening of the AVI has no effect on the next retrograde event (V-P2 is the same as V-P1). This is defined by the device as ELT, and a termination algorithm is activated. System is used by ELA Medical (Sorin Group, Milan).