Published ahead of print on August 5, 2004, doi:10.1164/rccm.200402-211UP
© 2004 American Thoracic Society doi: 10.1164/rccm.200402-211UP
Permanent Pacemakers and Implantable DefibrillatorsConsiderations for IntensivistsDepartment of Internal Medicine, Sections of Cardiology and Pulmonary/Critical Care, Bridgeport Hospital and Yale University School of Medicine, Bridgeport, Connecticut Correspondence and requests for reprints should be addressed to Constantine A. Manthous, M.D., Bridgeport Hospital, 267 Grant Street, Bridgeport, CT 06610. E-mail: pcmant{at}bpthosp.org ABSTRACT Pacemakers and cardioverter-defibrillators are implanted in patients with cardiovascular disease for an ever-increasing array of indications. Intensivists provide care frequently for patients who have these devices, and thus, they must be familiar with common problems and nuances that may contribute to critical illness. Close collaboration of the critical care physician and cardiologist/electrophysiologist assures that pacemakers and defibrillators are tuned to optimize the hemodynamic milieu of critically ill patients. Many recent advances in the sophistication of implanted devices are reviewed herein.
Key Words: arrhythmia • critical care • implantable cardiac defibrillator • pacemaker A permanent pacemaker (PPM) was first implanted in 1958 (1) and the first implantable cardioverter-defibrillator (ICD) in 1980 (2). In the ensuing years, as technologic developments have simplified implantation and expanded indications for their use, these devices have been placed in hundreds of thousands of individuals. Critical care physicians care for many patients who have these devices. Although most intensivists do not analyze and program these devices, it is important that they understand basic device function and how PPMs and ICDs may impact illnesses of their patients. Table 1 lists five questions for intensivists as they care for patients with PPMs and ICDs.
QUESTION 1: IS A DEVICE PRESENT, AND IF SO, WHAT IS IT? An infraclavicular (occasionally upper abdominal) scar with a palpable generator beneath may inform the examiner that a device is present. In obese patients or those in whom the device is implanted beneath the pectoral muscle, it can be hard to palpate modern, small PPM generators (or distinguish a PPM from an ICD). The chest radiograph illuminates the type of device (Figure 1). Patients and/or their family members usually carry wallet cards with essential information about their brand and settings of PPM or ICD. As a general rule, devices should be interrogated in the first 24 to 48 hours of admission whenever the PPM and/or ICD may contribute to illness.
QUESTION 2: MIGHT THE DEVICE BE CONTRIBUTING TO THE PATIENT'S PROBLEM? Intensivists should be familiar with basic PPM functions. Figures 2– 4 summarize common pacing modes. A detailed review of PPM function and electrocardiogram interpretation is beyond the scope of this article. Interested readers are referred to Internet resources, cited at the end of this article, that provide excellent reviews of PPM function, historical development, electrocardiogram patterns of PPM function and malfunction, and detailed discussions of indications for PPM and ICD implantation.
Malfunction of an implanted device may contribute to a patient's acute decompensation. This may take several forms. Battery depletion causes PPMs to reprogram automatically to slower rates, single-chamber pacing, and/or nonphysiologic (fixed rate) pacing modes. All of these changes may diminish cardiac output, the importance of which is discussed later here (after question 4). Lead fracture is diagnosed by a marked increase in lead impedance and/or visible lead discontinuity on chest radiograph. Fracture of a pacing lead may cause capture failure (electrocardiogram demonstrating pacing stimuli with no following P-wave or QRS) and bradycardia (an example of this is depicted in Figure E1 of the online supplement). Fracture of an ICD lead may cause defibrillation failure, leading to syncope, cardiac arrest, or death. It should also be noted that not all capture failure is related to lead malfunction. Marked hypokalemia, hypomagnesemia, and hyperkalemia may also contribute. Also, antiarrhythmic drugs such as flecainide and propafenone may increase the energy required to capture the myocardium. Lead insulation break is diagnosed by reduced lead impedance and is usually not visible on a chest radiograph. It results in two potential problems. First, during pacing, current may leak out the lead body, thereby reducing current that is delivered to the lead tip. This may result in capture failure. Second, and more commonly, the insulation break allows the device to sense chest wall myopotentials (or "noise") in addition to cardiac signals. In an atrial lead, this will inhibit atrial pacing with potential loss of the atrial contribution to cardiac output. In a dual-chamber PPM, the ventricular lead will track the noise on the atrial lead, resulting in ventricular pacing at the programmed upper rate of the device (an example of this is depicted in Figure E2 of the online supplement). Alternatively, if "mode switching" is programmed on, when the ventricular lead senses a rapid atrial rate, the device may switch to single-chamber pacing, resulting in loss of the atrial contribution to cardiac output (an example of this is depicted in Figure E3 of the online supplement). In a ventricular lead, noise will inhibit pacing output, resulting in pauses and bradycardia that can lead to syncope. Noise in the ventricular channel of an ICD is interpreted as ventricular arrhythmia and may prompt inappropriate delivery of pacing or shock therapy. Leads may also develop sensing malfunction unrelated to insulation breaks. Undersensing may be caused by fibrosis at the lead tip, injury of adjacent myocardium, or improper programming of the device's sensing parameters (an example of lead failure likely related to infarction of myocardium at the electrode tip is depicted in Figure E4 of the online supplement). Atrial undersensing may result in atrial pacing too soon after spontaneous atrial activity (P waves). In susceptible patients, this may induce supraventricular tachycardia, atrial flutter, or atrial fibrillation. In analogous fashion, ventricular undersensing may induce ventricular tachycardia. Undersensing of ventricular activity by an ICD may lead to failure to detect arrhythmia, syncope, cardiac arrest, and death. Oversensing of cardiac signals may result from improper programming of sensing settings or chamber cross-talk (detection of activity in one chamber by a lead in the other). In pacing systems, this usually leads to inhibition of pacing output (pauses or bradycardia). In dual-chamber pacing, however, oversensing by the atrial lead may lead to ventricular pacing at the upper rate or mode switching (discussed previously here). Oversensing by an ICD may prompt inappropriate delivery of therapy. A properly functioning device may be improperly programmed. The lower rate may be too slow for the patient's physiologic needs, especially in the critical care setting where sepsis, hypoxia, hypotension, and related conditions commonly demand faster resting heart rates. Failure to program properly the atrial sensing interval that follows a QRS (termed the postventricular atrial refractory period) may result in pacemaker-mediated tachycardia. After a sensed or paced ventricular depolarization, retrograde conduction to the atrium may be sensed there and relayed back to the ventricular lead, which paces, resulting in another retrograde atrial impulse and the cycle repeats. This results in ventricular pacing at or near the programmed upper rate of the PPM. The consequent periods of rapid ventricular pacing may be mistaken on monitor for ventricular tachyarrhythmias and may precipitate coronary ischemia and/or ventricular dysfunction in susceptible individuals. Finally, evidence has recently mounted (3, 4) to suggest that persistent right ventricular pacing may cause heart failure due to ventricular dyssynchrony from iatrogenic left bundle branch block. Recent data suggest that right ventricular pacing for as little as 20% of total beats may lead to left ventricular dysfunction in some patients (4). Patients presenting with congestive heart failure and frequent right ventricular pacing may benefit from reprogramming of the atrioventricular (AV) interval to minimize right ventricular pacing. Alternatively, they may require upgrade to biventricular pacing. Given their programming complexity, modern pacing systems often create diagnostic confusion even when they are functioning normally. Unexpected heart rate increases may result from sensor-mediated pacing (especially when the PPM is programmed to increase the paced rate in response to physiologic signals other than activity such as minute ventilation, body temperature, and pH). Atrial overdrive suppression is a feature in which the atrial rate is programmed to increase briefly after a premature atrial contraction or pause to suppress the occurrence of bradycardia-dependent atrial arrhythmias. One study demonstrated that such pacing may reduce atrial arrhythmia burden by 25% in some patients (5). Fast ventricular pacing may represent ventricular tracking of atrial arrhythmia or, in an ICD, antitachycardia pacing designed to terminate ventricular tachycardia. An unexpectedly slow heart rate may occur. Some PPMs have a programmable sleep mode, in which the device is programmed to a slower rate during nighttime hours. If unrecognized, this behavior may prompt a call from the night nurse that the PPM has stopped pacing.
Device Infection Intravenous catheter-related and other causes of bacteremia are quite prevalent in critically ill patients. The precise risk for secondary device infections after bacteremia is not entirely clear, although the rate of device infections in patients with staphylococcus bacteremia was 28% or more in one study (6). There are no explicit consensus guidelines to guide the duration of antibiotics in patients with bacteremia and no demonstrable involvement of device leads. It is generally believed that antibiotic therapy for bacteremia is also effective in treating microscopic involvement of the device (if it is present), although this has not been studied in a prospective manner. Accordingly, after cessation of antibiotics for primary bacteremia in which device infection is not suspected or proven, clinicians should follow patients carefully for signs and symptoms of recrudescent infection. QUESTION 3: MIGHT THE DEVICE BE PART OF THE DIAGNOSTIC SOLUTION? Patients with implantable devices provide unique diagnostic opportunities. Examination of intracardiac electrograms, obtained by interrogating the device with the manufacturer-specific programmer, may allow diagnosis of spontaneous arrhythmia mechanisms that are not apparent on surface electrocardiogram. Patients with an ICD who suffer syncope should always have the device interrogated to determine whether a ventricular tachyarrhythmia caused the event. During rapid ventricular tachycardia (VT), syncope may occur before ICD discharge, and thus, the patient will not report experiencing a shock (an example of this is depicted in Figure E5 of the online supplement). Such patients must not drive a motor vehicle until their syncopal VT is better controlled. Also, modern devices store heart histogram data that provide insight into chronotropic competence, heart rate in atrial fibrillation, and spontaneous arrhythmia burden. Arrhythmia logs (often with stored electrograms) detail past spontaneous arrhythmias. Data from one recent PPM trial that assessed utility of single versus dual-chamber pacing revealed that the majority of first strokes occurred in patients whose arrhythmia logs demonstrated frequent periods of asymptomatic atrial fibrillation flutter (7). Such knowledge gleaned from PPM interrogation may delineate the source of stroke or help to identify patients who may benefit from systemic anticoagulation therapy before they present with a first stroke (an example of this is depicted in Figure E6 of the online supplement). Finally, interrogation of the device will help exclude device malfunction or improper programming as detailed previously here. QUESTION 4: MIGHT THE DEVICE BE PART OF THE THERAPEUTIC SOLUTION? From the foregoing discussion, it is apparent that many of the problems to which a device may be contributing may be corrected by careful reprogramming of the device. The lower rate may be increased to improve cardiac output and potentially suppress bradycardia-dependent arrhythmias. One example of the latter is atrial fibrillation/flutter, which in some patients occurs during periods of vagal-mediated bradycardia. Another example is Torsades de pointes. When this potentially fatal arrhythmia occurs in a patient with a device, increasing the lower rate to 80 or 90 per minute may suppress its occurrence. This may temporize, allowing time to correct an underlying malady that is contributing to genesis of the long corrected QT interval that promotes the ventricular tachycardia. The AV interval may also be altered to improve cardiac output. In some cases, it is appropriate to lengthen this interval to promote intrinsic conduction and minimize right ventricular pacing that may contribute to heart failure. In other circumstances, it may be possible to "optimize" the AV interval by performing echo-Doppler studies to select an AV interval that optimizes left ventricular filling and emptying (8). In some critical care situations, reprogramming of the pacing mode may be appropriate. If tracking of atrial arrhythmias is causing inappropriately rapid pacing, reprogramming to single-chamber mode will prevent this until the arrhythmia can be controlled. In the rare patient with a dual-chamber PPM programmed to single-chamber ventricular pacing, reprogramming to dual-chamber pacing restores AV synchrony. Patients with a rate-adaptive PPM that tracks physiologic parameters (e.g., minute ventilation) may experience inappropriately rapid pacing, which can be prevented by deactivating the rate-adaptive function. Not infrequently, patients present with shock, the primary pathogenesis of which is not related to arrhythmia. If the PPM does not track physiologic parameters and is set at a lower rate that is too slow for stress states, an inappropriately low cardiac output may contribute to shock. In these situations, the intensivist and cardiologist can work carefully together to adjust the paced rate, using physiologic parameters (e.g., blood pressure, urine output, lactic acid levels, and pressure/output measurements when a pulmonary artery catheter is present) to determine optimal settings. There will be a critical heart rate after which further increments reduce cardiac output and blood pressure caused by insufficient diastolic filling time and/or development of myocardial ischemia in vulnerable individuals. Optimal settings are likely to change from hour to hour as myocardial mechanical loading and underlying shock evolve (9, 10). Accordingly, frequent reassessment of the "best" heart rate is likely to minimize the degree to which insufficient cardiac output contributes to systemic underperfusion. Finally, implanted devices may be used to terminate spontaneous arrhythmias. Using the PPM programmer, atrial burst stimulation may be delivered to terminate atrial tachycardia or flutter. Ventricular burst stimulation may be delivered through a permanent pacer to terminate organized ventricular tachycardia (although this may prompt ventricular fibrillation so an external defibrillator must be at the ready). Through an ICD, a commanded discharge may be delivered to terminate ventricular or atrial arrhythmias. QUESTION 5: IS THE PATIENT A CANDIDATE FOR DEVICE IMPLANTATION? The ever-expanding indications for PPM and ICD implantation are beyond the scope of this article. The reader is referred to recently updated guidelines published jointly by the American Heart Association and the American College of Cardiology, available in print (11, 12) and on the Internet (see Internet references and Table 2 for "Class 1" indications). The therapeutic need for permanent PPM (e.g., Mobitz II AV block, complete heart block, symptomatic sinus node dysfunction) and ICD (sustained ventricular arrhythmias) is often obvious. It may be difficult to determine, however, when arrhythmias in the acute setting are primary problems or secondary to metabolic perturbations such as ischemia, hypoxia, electrolyte abnormalities, arrhythmogenic medications, and hyperadrenergic states. Often in such situations, the best approach is to correct metabolic abnormalities, use antiarrhythmic drugs to suppress acute arrhythmias, and perform electrophysiologic evaluation once the patient is stable.
Patients with systolic dysfunction and congestive heart failure despite medical therapy may benefit from biventricular pacing (cardiac resynchronization therapy). Although present guidelines recommend this therapy only for patients with QRS of more than 120 milliseconds, data are accumulating to suggest that the electrocardiogram is not the optimal tool by which to select appropriate candidates for such pacing. Tissue Doppler imaging and related echocardiographic technologies permit demonstration of ventricular dyssynchrony. This noninvasive imaging may predict a good response to biventricular pacing (13, 14) even in some patients with a QRS duration of less than 120 milliseconds. (15). A recent editorial provides an excellent summary of this issue (16). A comprehensive review of the topic is also available (17). Presently, prophylactic ICD implantation is indicated in survivors of myocardial infarction with spontaneous nonsustained VT and left ventricular ejection fractions of more than 30% but 40% or less when sustained ventricular tachycardia can be induced during an electrophysiology study. Survivors of myocardial infarction with left ventricular ejection fractions of 30% or less may receive an ICD without preliminary electrophysiology study if their QRS duration is more than 120 milliseconds (11). Mounting data suggest that the QRS duration may be too restrictive. Also, recently published studies have demonstrated that prophylactic ICD implantation improves survival in individuals with reduced left ventricular ejection fraction in the context of nonischemic cardiomyopathy (18–20). Although these findings have not yet been incorporated into published implant guidelines or payment reimbursement criteria, it is anticipated that they soon will be so included. In general, electrophysiologic evaluation of patients considered for prophylactic ICD implant should be deferred until acute illness has subsided. In fact, present guidelines dictate that such studies not be performed less than 30 days after myocardial infarction or less than 90 days after coronary revascularization. Nonetheless, it is appropriate to seek electrophysiology consultation during the acute illness to assist with acute arrhythmia management and plan future study.
Special Considerations
External cardioversion and defibrillation.
Interaction of devices with magnets.
Electromagnetic interference.
Device considerations at the end of life. CONCLUSION The foregoing discussion highlights many of the complexities of modern PPM and ICD devices. These devices play an increasingly important role in the management of critically ill patients. Although critical care physicians are not responsible for implantation, analysis, or programming of these devices, it is incumbent that they identify patients with PPM and ICD units, consider the questions outlined in Table 1, and request that the implanted device be interrogated shortly after hospital admission. Interrogation may be performed by the cardiologist who follows the patient's device, the electrophysiology-pacemaker service of the hospital, or a representative of the device's manufacturer. Selected patients may develop indications for device placement. Cardiologists and intensivists should work collaboratively to define solutions for critically ill patients with arrhythmia and/or antiarrhythmia devices. APPENDIX:
INTERNET REFERENCES AND RESOURCES (All accessed: February 19, 2004)
FOOTNOTES This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org Conflict of Interest Statement: C.A.M. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; C.M. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form February 19, 2004; accepted in final form August 4, 2004 REFERENCES
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
