Many patients in the outpatient cardiology clinics, electrophysiology clinics and geriatric clinics have permanent pacemakers. These devices are mainly fifth-generation pacemakers with electrogram storage of arrhythmic events. Background A routine clinical pacemaker interrogation includes measurement of battery voltage, pacing threshold, impedance, and sensing functions. An important feature that is less well known is the arrhythmia monitoring and recording feature of the pacemakers. Most pacemakers can be programmed to record high-rate atrial and ventricular events; their results may have significant bearing on patient management. For example, multiple episodes of atrial fibrillation with prolonged duration or rapid ventricular response suggest consideration for anticoagulation and rate control. Episodes of possible ventricular tachycardia can be documented for diagnostic and therapeutic purposes. Methods Examples of the problems faced by the clinician during analysis of stored pacemaker data are shown in Figures 1 and 2, as well as Table 1. Figure 1 shows an example of a stored composite electrogram with both atrial and ventricular electrograms during an episode of tachycardia. The result is that a series of atrial electrograms (marked by arrows) are not sensed. The pacemaker software may then identify an episode of ventricular tachycardia. Inspection of the tracings shows the correct diagnosis to be supraventricular tachycardia initiated by an atrial complex and continuing with an atrial depolarization before every ventricular depolarization. The pacemaker report may label the episode incorrectly as ventricular tachycardia. The important lesson from this tracing is that careful inspection of stored electrograms from pacemaker interrogation may be required to make the diagnosis. Atrial depolarizations occurring later in the cardiac cycle are sensed as refractory events and are counted for diagnostic purposes (AR), but do not initiate a ventricular paced complex. The pacemaker initiates a programmable period of altered atrial sensing to avoid oversensing of electrograms on the atrial channel. The PVARP (Post Ventricular Atrial Refractory Period) is programmable (approximately 150â€“500 msec). The first part of the PVARP may be programmed to avoid any sensing of the atrial electrogram Post Ventricular Atrial Blanking Period resulting in the loss of detection of atrial electrograms (as shown in Figure 1). Figure 2 shows an example of a tracing obtained from a more recent pacemaker implant. The pacemaker is programmed to sense atrial depolarizations at all times so that atrial electrograms are detected (see arrows) throughout the post ventricular atrial blanking period. This algorithm is similar to that used in defibrillators. Note that a ventricular depolarization is sensed on the atrial electrode (labeled Ab, as well as an atrial depolarization in the post ventricular atrial blanking period, also labeled Ab). This is an example of far-field oversensing of the ventricular electrogram on the atrial channel. If far-field sensing is frequent, an inappropriate diagnosis of atrial fibrillation can be made with the potential for improper therapy. Review of the electrograms in Figure 2 shows that all other atrial depolarizations are sensed and labeled properly on the marker channel. Note that two atrial electrograms are present with one ventricular depolarization in the middle of the tracing, confirming the diagnosis of supraventricular tachycardia. The older pacemaker software acts to prevent oversensing of a ventricular depolarization by providing a post ventricular atrial blanking period (no atrial sensing) after the ventricular depolarization, followed by a post ventricular atrial refractory period where atrial complexes are sensed, but do not initiate a subsequent ventricular paced complex. Therefore, pacemaker-mediated tachycardia is prevented, since atrial depolarization in the PVARP does not result in a paced ventricular complex. As shown in Table 1, software improvements consist of expanded memory for atrial and ventricular events and sensing of atrial depolarizations throughout the cardiac cycle. As seen in Figure 2, the increased time for sensing may result in far-field sensing of the atrial electrogram in some patients. The absolute values of each parameter will vary between pacemaker vendors. The most appropriate recommendation is to base treatment on all available data including a careful inspection of stored electrograms. Discussion Improvements in pacemaker memory and arrhythmia detection have been confirmed by the study of electrograms and comparison with surface ECG or Holter recordings.2â€“5 The recording of both atrial and ventricular electrograms allows examination of both pacemaker tabulations of supraventricular and ventricular arrhythmias as well as direct examination of stored electrograms. The goals of rhythm analysis by clinical examination, ECG, or stored data from pacemakers or other devices include detection of arrhythmias (e.g., atrial fibrillation, atrial flutter, or ventricular tachycardia) that may alter therapy. In the patient with a permanent pacemaker, the initial step is to examine all available clinical data to be certain that the patient does not have ventricular pacing with unrecognized atrial fibrillation. A comparison of the frequency of correct diagnosis of atrial fibrillation in patients with and without continuous ventricular pacing has demonstrated that the diagnosis of atrial fibrillation and optimal use of anticoagulation are significantly lower in patients with continuous ventricular pacing.6 If the atrial rhythm is uncertain in a patient with dual-chamber pacing, direct examination of the atrial electrograms may not provide the diagnosis. After determination of the baseline rhythm, the stored electrograms from all arrhythmias should be examined and compared with the pacemaker tabulation. In current devices, the relationship of ventricular and atrial electrograms during arrhythmias can be determined (Figures 1 and 2), and a presumptive diagnosis made. Although the automated detection of ventricular and supraventricular arrhythmias may be improving, significant problems remain. Difficulties include the diagnosis of brief episodes of supraventricular tachycardia,7 false positives due to far-field sensing and myopotential sensing,8 and changes in voltage of the atrial electrogram during atrial arrhythmias.9 Comparison of single- and dual-chamber arrhythmia detection by cardioverter-defibrillators10 has shown a relatively high misclassification rate for episodes of atrial tachycardia with 1:1 conduction (98% incorrectly classified) and atrial flutter (50%). The correct diagnosis can often be made by careful comparison of stored electrograms with the pacemaker marker channel. Diagnosis of unsuspected atrial fibrillation by stored pacemaker electrograms may influence clinical care and patient outcomes. The incidence of atrial fibrillation increases with age and may be asymptomatic. The patient may not describe palpitations, yet be at risk for embolic stroke. The rapid ventricular response may increase myocardial oxygen demands and result in angina, tachycardia-related cardiomyopathy, and congestive heart failure. Recognition of the patient at risk may permit a change in anticoagulation strategy and improved rate control. Patients with coronary artery disease may have ventricular arrhythmias related to both coronary ischemia and myocardial scars that predispose to slow conduction and reentry. Conclusions Increased storage of atrial and ventricular electrograms obtained during continuous monitoring in permanent pacemakers allows confirmation of significant arrhythmias by direct examination of the stored data. The major considerations for the clinician are anticoagulation and rate control for previously unrecognized atrial fibrillation and evaluation of ventricular tachycardia for device or drug therapy. Therefore, the stored pacemaker electrograms serve as a silent witness to atrial and ventricular arrhythmias.