When Your Patient Has an Implantable Cardioverter-Defibrillator

Emerg Med 39(03):30-37, 2007

It's a double-edged sword, a device that can save a life but can also break down in any number of ways. Here's an update on how ICDs work—and how they can fail.

By Charles S. Graffeo, MD, and James D. Krygowski, MD

Dr. Graffeo is a professor and assistant residency director and Dr. Krygowski is a senior resident in the department of emergency medicine at the Eastern Virginia Medical School in Norfolk.

According to the latest statistics, implantable cardioverter-defibrillators (ICDs) are being placed into 100,000 patients in the United States annually. In fact, ICDs are becoming the standard of care in managing recovery from myocardial infarction when the patient has both ventricular tachyarrhythmias and reduced ejection fraction. Recent data suggest that ICDs are superior to drug therapy for arrhythmia management in this population. Another indication for ICDs is cardiac resynchronization therapy in patients with class III or IV congestive heart failure.

In general, the majority of ICD patients encountered in the emergency department are elderly and typically have a history of advanced cardiac disease. There is often dysfunction of the myocardial muscle, electrical conduction system, or cardiac vessels. In addition, these patients frequently have associated comorbid conditions such as diabetes or renal insufficiency. On occasion, a young patient with an ICD may be seen, as in the case of individuals with Brugada syndrome, long QT syndrome, or cardiomyopathy (hereditary or post-pregnancy).

In many instances, ICDs can be credited with extending the lives of patients who otherwise would have succumbed to a fatal arrhythmia. At the same time, it is clear that ICD placement is not without risk, as the device can malfunction in a variety of ways. Thus, it is important for emergency physicians to be able to recognize ICD breakdowns and properly manage patients prior to cardiac consultation (ideally with an expert in electrophysiology) and to know which patients may be safely discharged from the emergency department and which patients require hospital admission.

purpose of an icd

The primary role of an ICD is to deliver electrical energy to the myocardium to abort ventricular tachycardia (VT) or ventricular fibrillation (VF). Most ICDs utilize a biphasic current, which reverses direction in mid-shock instead of traveling in one direction only, as was the case with the first generation of ICDs. Multiple studies have shown that biphasic defibrillation is significantly superior to monophasic defibrillation, requiring less energy and causing less damage with greater success rates.

Many of the new ICDs have antitachycardia pacing. With this feature, the device will send a series of extra-fast pacing beats in an attempt to restore sinus rhythm before discharging a full shock. If this is not successful, the device reverts to a standard high-energy shock that is typically between 1 and 50 joules. In addition, ICDs have an antibradycardia function that works like a pacemaker. If the heart rate drops below a certain rate, the ICD will provide a pacing function with significantly less forceful shocks.

There are three main components to an ICD (see image below). The first is a small computer that is able to analyze and record the patient’s heart rate and rhythm. This computer can be queried as to the device history (such as rhythms and interventions) and program settings, which can be changed. The second component is the wires (or leads) that are usually placed through a vein (most often the left subclavian vein) into the right ventricle for single-chamber devices or both the right atrium and ventricle for dual-chamber devices. The third component is the battery that powers the ICD, which usually lasts 5 to 10 years, depending on the settings and usage.

Components of an ICD. This posteroanterior chest radiograph shows a pacemaker pulse generator and leads. The small arrow indicates the atrial pacing lead. The large arrow indicates the ventricular pacing lead.

Identifying an icd

Most patients with an ICD are well aware that they have the device and may in fact be able to provide you with an information card that usually includes the type of device, manufacturer, model numbers, date of placement, electrophysiologist, and contact telephone numbers. In the event that an aphasic or unconscious patient presents to an emergency department with an ICD in place, the most likely location for the device is in the left upper chest, where it may be palpated on physical examination. However, it should be noted that it is becoming increasingly difficult to palpate these devices because of their ever-diminishing size. The latest ICD dimensions are only about 30 to 60 cc in volume, and in an obese patient this may not be detectable. In some cases, a scar on the left anterior chest wall may be the best clue to the presence of an ICD.

A chest x-ray is an excellent way to visualize an ICD. If an information card is not available, it is possible to shoot an over-penetrated, close-up x-ray directly at the device. Manufacturers are required to mark ICDs with symbols or numbers that are visible on radiographs identifying the company name as well as the specific model of the device. This information can be crucial in a noncommunicative patient because each ICD manufacturer has a specific way to interact with a particular model. (Currently, the three major ICD manufacturers are Medtronic, Guidant, and St. Jude.) Usually, a device representative will bring a special computer with a magnetic wand to place over the ICD. The computer will then “talk” to the ICD, obtaining information such as settings, rhythm history, and shock history. The settings can be adjusted if needed.

setting thresholds

A cardiologist will set the thresholds that allow an ICD to recognize VT or VF. There are a number of variables that can be adjusted to change the device’s sensitivity and specificity for these arrhythmias. The different options depend on whether the ICD is a single-chamber or dual-chamber device.

In general, the ICD must first recognize that the patient’s heart rate is “fast.” The cardiologist will set the threshold rate, and the device will recognize any rate above that as tachycardia. Cardiologists can also use rate-altering medications to allow them to choose a lower threshold rate. Both beta blockers and calcium channel blockers slow the heart rate, making it less likely that a patient will be shocked while in atrial fibrillation or a sinus rhythm.

The cardiologist will also set a “suddenness” variable to avoid shocks in patients who gradually increase their heart rate during exercise. The vast majority of VT/VF events are of sudden onset, so this type of programming makes sense and is most welcome in cardiac rehabilitation patients who would otherwise have a rude surprise soon after beginning to exercise.

Another adjustment that is available involves the “regularness” of the rhythm. An ICD can be programmed to detect atrial fibrillation (AF) and yet withhold defibrillation because it recognizes the rhythm to be regular. While it has been proposed that ICDs could be used to convert atrial arrhythmias, there are insufficient data to warrant this as a standard of care.

Many ICDs are able to compare the vector morphology on the patient’s ECG to his preexisting sinus morphology, which often helps to distinguish VT from rhythms originating above the ventricles. The device will record the appearance of the patient’s normal ECG while in sinus rhythm. A new “fast” morphology will be compared with this stored rhythm, and if it is uniform and identical, it is classified as supraventricular tachycardia (SVT) and firing is withheld. Alternatively, if the new rhythm is uniform and distinctly different from the preexisting sinus morphology, it is classified as VT and firing commences.

advantages of dual-chamber devices

With dual-chamber ICDs, there is a lead in both the right atrium and ventricle, which offers several advantages. If a dual-chamber device senses that the ventricular rate is faster than the atrial rate, the device will classify the arrhythmia as ventricular in origin and fire, assuming the heart rate is above the threshold rate. If it senses that the atrial rate is faster, then firing will be withheld because the ventricular rate is most likely the result of AF or an SVT. If the two rates are the same, the ICD will compare the morphology to normal sinus rhythm to determine if a shock is needed.

Dual-chamber ICDs are particularly helpful in patients with cardiomyopathies, who frequently go into AF with rapid ventricular response. This can be very difficult for a single-chamber device to differentiate.

Patients with ICDs who present to the emergency department for evaluation experience shocks that are typically very painful, even though significantly less energy is used compared to external defibrillation. (The shocks are often described as feeling like a “swift kick in the chest.”) As noted earlier, shocks usually run from 1 to 50 joules, depending on the patient’s defibrillation threshold. This threshold is determined by a variety of factors and is measured and tested at the time the device is implanted.

The direction of the charge is programmed by the electrophysiologist, but generally it travels from the distal ventricular lead (located in the right ventricle) to the casing of the ICD. In many models, the manufacturer places a second coil in the ventricular lead but more proximal to the ICD, at the level of the superior vena cava. In such cases, the charge will travel from the distal ventricular lead to both the ICD casing and the second coil. In either case, whether the ICD has a second coil or not, the direction of the charge is reversed mid-shock if the device uses biphasic energy.

Patients who experience defibrillation while awake are generally anxious and fearful of refiring. Depression is also common in these patients because the firing of their ICD sometimes serves to remind them of their heart disease.

We noted earlier that if the patient’s heart rate drops below a certain set rate, then the ICD will function as a pacemaker. This can take a significant toll on the battery. For this reason, the cardiologist will be very careful in choosing settings. They need to be patient-specific. A patient who remains active, frequently rises from a sitting to a standing position, and lives alone will have very different settings from those chosen for a nonambulatory nursing home patient.

three types of shocks

Patient complaints associated with ICDs include the shocks themselves, palpitations, syncope, chest pain, signs of infection, and beeping. Other situations unique to ICDs are post-placement events such as lead failure, twiddler syndrome, and treatment of cardiac arrest. There are a number of other miscellaneous issues that may be very important to an emergency physician, such as the use of magnets, magnetic resonance imaging (MRI), cell phones, and the placement of central lines.

The most concerning (and device-specific) complaints are the shocks that ICD patients experience. In the broadest sense, there are three types of shocks: appropriate, inappropriate, and phantom (see table below). These are the qualifiers that cardiologists will use to determine what happened to a patient who was shocked. This is not always possible to determine while the patient is in the emergency department, and in view of this it is important for the evaluating physician to distinguish between shocks that are highly concerning and those that are concerning. In any patient with an ICD that has fired, it is generally recommended that the patient’s electrophysiologist be contacted while a workup is under way. In cases where rapid redundant firing occurs, emergent cardiology consultation (preferably with an expert in electrophysiology and ICDs) is indicated.

Appropriate shocks. These shocks occur when an ICD accurately senses VT/VF and delivers a high-energy shock to the heart in an attempt to convert the arrhythmia to a normal sinus rhythm. These shocks are noted to be appropriate either when device interrogation confirms the presence of VT/VF or when the patient is placed on a cardiac monitor while in the emergency department and is found to be in VT/VF and the device fires. In such cases, it is imperative to determine whether the patient is having an acute ische-mic event or other precipitating events that predispose to arrhythmias. (Of note, it is not uncommon for patients to have a general sense of ill-being prior to being shocked by an ICD.)

Inappropriate shocks. When an ICD is fooled into believing the patient is in VT/VF and, following its programmed algorithm, delivers a shock, it is considered an inappropriate shock. There are many ways this can happen, and it is difficult to determine the etiology in the emergency department. One common scenario is that the patient’s normal sinus rhythm or AF incites a rapid ventricular response, which the ICD misinterprets as a ventricular tachycardia. As programmed, the ICD firing usually converts AF, which can be discovered on querying the device. The patient may or may not have realized he was experiencing AF, but he will almost certainly report the pain caused by the ICD shock.

Another possibility is that the ICD misreads either a T wave or P wave and interprets it to be a QRS complex, effectively doubling the actual heart rate. R waves can also be counted twice by an ICD, which can cause the device to fire. Although it is far less common now, some patients may have both a pacemaker and an ICD. This extra electrical cardiac activity can make it difficult for the ICD to read and accurately differentiate a normal paced rhythm from an arrhythmia.

Lead failure can occur with inappropriate shocks and can also result in failure to shock. Additionally, diaphragmatic or pectoral muscles may create myopotentials that can be mistaken for heart beats. Finally, external electromagnetic interference (from cell phones or MRI, for example) can interfere with an ICD and cause inappropriate shocks.

Phantom shocks. These are perceived shocks that are not associated with the discharge of energy from an ICD. They can only be diagnosed by a device-specific communicator. It may be useful to ask the patient if the shock felt like a “big” shock or a “small” shock. It is very unlikely that a patient will attribute real defibrillation to a small shock; it is more likely that the patient experienced antitachycardia pacing, not defibrillation. However, consultation with an electrophysiologist should still take place and the device should be queried.

Highly concerning shocks. Repetitive shocks or those that occur more than three times in 24 hours are considered highly concerning. Under no circumstances should these patients be discharged home before the device has been interrogated. It is very likely that either the patient has a new underlying condition, prompting VT/VF, or the device has a serious defect. The patient may have an acute ischemic event, serious electrolyte abnormality, or drug interaction. Class I antiarrhythmic drugs and amiodarone may increase the amount of energy required to successfully defibrillate the heart. The lead on the device may have dislodged, which may be visible on x-ray, or it may have fractured, which cannot be seen on x-ray. Both mishaps typically render the ICD nonfunctional.

For these reasons, any patient with highly concerning shocks should be considered to have a malfunctioning ICD until proven otherwise. These patients would all require telemetry monitoring with defibrillator pads, an ECG, chest x-ray, basic metabolic profile, and cardiac enzyme levels (at a minimum).

Concerning shocks. These are isolated events in a well-appearing, hemodynamically stable patient. In such cases, after a screening ECG and a telephone consult with the electrophysiologist, it is generally recommended that the patient be reassured and discharged home. The initial workup of this type of patient is at the discretion of the emergency physician, but it is within the standard of care to send this patient home without laboratory work or imaging, provided there has only been one isolated shock, the patient looks well, his electrophysiologist has been consulted, and all parties involved agree with the plan.

There are well-recognized psychological responses to shocks. Patients commonly have symptoms of anxiety, such as sweating and increased heart and respiratory rates. Risk factors for anxiety in patients with an ICD include a relatively young age and increased frequency of shocks. Although it is difficult to address in the emergency department, it may be beneficial to promote relaxation strategies, including deep breathing, positive thinking, prayer, or meditation. Ideally, the cardiologist has already counseled the patient and they have developed an action plan for possible shocks, but this is not always the case.

other presenting complaints

Besides shocks, other presenting complaints in patients with an ICD include palpitations, syncope, chest pain, signs of infection, and beeping. We will now discuss each of these in turn.

Palpitations. Arrhythmias, of course, are common in patients with ICDs. The patient should be asked whether his heart rate feels fast, slow, regular, or irregular. Recent medication changes, caffeine use, and other medical problems may precipitate palpitations. A standard ECG, telemetry, and possibly rhythm strips might be helpful in interpreting the rhythm. The recording ability of each ICD device varies, with higher-end models capable of capturing full electrocardiograms, while others only include a time log with each arrhythmia. The need for further monitoring, such as with a Holter monitor, should be discussed with the electrophysiologist.

Syncope. Syncope can be a sign of true VT/VF that has been cardioverted, especially if the patient is elderly. Standard measures should include telemetry with defibrillator pads, an ECG, chest x-ray, and electrolyte and cardiac enzyme levels. A computed tomography (CT) scan of the head should be considered, especially if trauma (such as from a fall) or anticoagulant medication is involved. The ICD should be interrogated while the patient is still in the emergency department, if possible, or soon thereafter. In general, these patients will be admitted to the hospital, although in some cases they may be discharged home directly from the emergency department after the ICD has been queried.

Chest pain. Chest pain should be treated aggressively in patients with ICDs. They should all receive standard-of-care diagnostic testing and treatment, including an ECG, telemetry, oxygen, intravenous access, aspirin administration, chest x-rays, and electrolyte and cardiac enzyme levels at a minimum. Patients with ICDs typically have ischemic cardiac disease and additional ischemic events should always be anticipated.

Signs of infection. Within the first six months of placement, an ICD, like other surgically implanted devices, can become infected with bacteria. The usual patient complaints in such cases are erythema and warmth at the placement site or fever. The cardiologist must evaluate these patients and will most likely remove the entire device, including the leads. Post-surgical tests such as a complete blood count, blood culture, and coagulation studies are usually indicated. Most patients will be started on antibiotics that cover common organisms, such as Staphylococcus aureus or Streptococcus pyogenes or, in high-risk patients, methicillin-resistant S. aureus.

Beeping. A patient may present to the emergency department because the device is emitting occasional beeping sounds. Although it may be easy to dismiss this as a potential psychiatric complaint or one of minor annoyance, the complaint of beeping requires a telephone consultation with the patient’s electrophysiologist. Often, the device beeps when its battery is running low. Usually, this is discovered during regular visits to the cardiologist, but with poor patient compliance or a sudden increased use of battery power (from pacing over long periods of time, for example), the battery may be drained. The patient should be placed on telemetry with defibrillator pads until the device is interrogated and the problem resolved.

acute post-placement issues

From minutes to days after the ICD is initially placed, a number of surgical complications can arise, including pericardial tamponade, cardiac perforation, air embolism, pneumothorax, venous thrombosis, pocket hematoma, and wound dehiscence. All of these should be considered in patients presenting to the emergency department in the immediate perioperative period.
Also, as noted earlier, various other post-placement issues may need to be addressed, as follows:

Lead failure. The leads of an ICD are sources of potential problems (see images below). Shortly after placement, a lead may perforate the endocardium, and over time leads have been known to dislodge or fracture. Additionally, the insulation of the lead may have a defect that can cause inappropriate shocks. Because lead dislodgement or fracture may not be visible on x-ray, detecting either malfunction or an insulation defect is extremely difficult in the emergency department. In addition, the patient’s symptoms can vary. Other variables, such as the material used for the lead or insulation, the design for the lead conductor, and the manufacturer, may play a role in a malfunction.
With any type of ICD malfunction, lead failure should be in the initial differential. Device interrogation is required to confirm the diagnosis.

Twiddler syndrome. Twiddler syndrome refers to patients who, consciously or unconsciously, twiddle with their ICD to the point that the leads dislodge, rendering the device ineffective. These patients are usually older women with lax subcutaneous tissue in which the ICD was placed. Obesity and psychiatric illness are other risk factors. Twiddler syndrome usually occurs within the first six months of device placement, so extra care should be taken when treating patients with recently implanted devices. Skin alterations at the placement site may be suggestive of this syndrome. In many cases, it is possible to visualize excessively twisted leads on a chest x-ray.

Cardiac arrest. Although it is true that an ICD should be delivering shocks, rendering an external defibrillator unnecessary, it cannot be assumed that the device is functioning properly should a patient go into cardiac arrest. An external defibrillator, therefore, should still be used just as if there were no ICD present, with one important exception. A paddle should never be placed directly on top of the ICD; this can damage the ICD and result in insufficient energy being delivered to the heart. Chest compressions may be performed as usual. Any shocks delivered by the ICD will not be painful to rescuers, especially if standard hospital gloves are worn. All other actions should be in accordance with advanced cardiac life support protocols.

Magnets. The use of magnets with ICDs is different than with pacemakers. Magnets placed directly over a pacemaker will suspend its function entirely. With an ICD, a magnet will suspend its tachyarrhythmia function but not its bradyarrhythmia function. In other words, the ICD will not deliver antitachycardia pacing or shocks in response to VT/VF, but it will continue to function as a pacemaker should the patient’s heart rate drop below the threshold rate. A magnet may be placed over an ICD when it is believed the device is delivering shocks inappropriately and firing needs to be terminated. All patients with an ICD who have a magnet placed on their chest should be on telemetry with defibrillation pads. It is best to tape the magnet to the chest (or directly over the ICD) so it will not fall off if the patient moves around.

Removing the magnet should allow the ICD’s tachyarrhythmia function to return, but the device should always be interrogated to confirm that the settings are what the cardiologist wanted. There have been a number of malpractice cases in which a patient was sent home after a magnet had been used on an ICD and the patient later died. When the ICD was interrogated, it was learned that the tachyarrhythmia function had been turned off by the magnet but had not returned when the magnet was removed. It is imperative, therefore, that the device be interrogated before the patient is sent home.

Magnetic resonance imaging. It has long been accepted that patients with ICDs are not candidates for MRI studies. The leads may be damaged or the device settings may be altered or even turned off. Although there have been some opinions expressed to the contrary, this is still the current standard of care.

Cell phones. Cell phones are not recommended for use in close proximity to an ICD. Like other forms of electromagnetic radiation, they could lead to inappropriate shocks or even turn off the device. There have been a few recent articles that have tested various ICDs in close proximity to cell phones without adverse interference errors. However, current risk-benefit analysis holds that it is prudent for ICD patients to avoid using or being around cell phones.

Central lines. Although there is no contraindication for the placement of a central line in patients with ICDs, caution is recommended. Catheter manipulation in the right heart can dislodge the ICD leads if they have not become fibrosed into the heart muscle, a process that usually requires a number of months. Also, when guidewires are used, it is advisable to avoid going into the heart because the ICD may sense it as extra electrical activity, which can cause the device to fire. If a guidewire contacts the ICD electrode during firing, it may damage the device.

new horizons

There are many innovative ideas being discussed and clinical studies under way regarding new applications for ICDs. A few animal studies have shown that when a lead is placed in the coronary sinus (middle coronary vein or great coronary vein), less energy is required to halt ventricular fibrillation. Ideally, this would be more tolerable to the patient and would permit longer battery life. However, such lead placements would be a more challenging procedure for the cardiologist.

New ICD leads have been proposed to monitor various measurements that may prove helpful to clinicians. A number of companies have developed pacing leads that are able to measure venous oxygen saturation, but the failure rate with these leads is high. Oxygen levels have been studied in patients with obstructive sleep apnea to evaluate apnea treatment. In the future, it may be possible to measure certain electrolyte levels directly from the heart with ultrafiltration or microdialysis techniques.

Using an ICD to terminate new-onset AF has been proposed as well. Benefits include immediate cardioversion and reducing the risk of stroke and other embolic events. The primary difficulty is patient tolerability; the shocks are painful without sedation. Considering how common AF is and the fact that it is relatively benign compared to VF, it is difficult to justify this new indication.

MAJOR IMPACT
Implanted cardioverter-defibrillators have had a major impact on the quality of life of many patients who may have otherwise succumbed to advanced cardiac disease. Their success and development have led to new challenges in understanding why various errors commonly occur. Given the dramatic increase in the number of patients with these devices, it is more important than ever for emergency physicians to be familiar with the latest advances in this state-of-the-art technology.

back to top

Suggested Reading Dosdall DJ and Ideker RE: New technologies of internal defibrillation. J Interv Card Electrophysiol 13(Suppl 1):67, 2005. Gradaus R, et al.: ICD leads: design and chronic dysfunctions. Pacing Clin Electrophysiol 26(2 Pt 1):649, 2003. Hood RE and Shorofsky SR: Management of arrhythmias in the emergency department. Cardiol Clin 24(1):125, vii, 2006. Kanz KG, et al.: Effect of digital cellular phones on tachyarrhythmia analysis of automated external defibrillators. Eur J Emerg Med 11(2):75, 2004. Kusumoto F and Goldschlager N: Beyond heart rhythms: new directions for implantable devices. J Interv Card Electrophysiol 14(1):5, 2005. Mirowski M, et al.: Termination of malignant ventricular arrhythmias with an implanted automatic defibrillator in human beings. N Engl J Med 303(6):322, 1980. Pinski SL: Emergencies related to implantable cardioverter-defibrillators. Crit Care Med 28(10 Suppl):N174, 2000. Rasmussen MJ, et al.: Unintentional deactivation of implantable cardioverter-defibrillators in health care settings. Mayo Clin Proc 77(8):855, 2002. Scher DL: Troubleshooting pacemakers and implantable cardioverter-defibrillators. Curr Opin Cardiol 19(1):36, 2004. Sears SF Jr, et al.: Cardiology patient page. How to respond to an implantable cardioverter-defibrillator shock. Circulation 111(23):e380, 2005. Sharifi M, et al.: Twiddling to the extreme: development of twiddler syndrome in an implanted cardioverter-defibrillator. J Invasive Cardiol 17(3):195, 2005. Stevenson WG, et al.: Clinical assessment and management of patients with implanted cardioverter-defibrillators presenting to nonelectrophysiologists. Circulation 110(25):3866, 2004. Swerdlow CD and Friedman PA: Advanced ICD troubleshooting: part I. Pacing Clin Electrophysiol 28(12):1322, 2005. Swerdlow CD and Zhang J: Implantable cardioverter-defibrillator shocks: a troubleshooting guide. Rev Cardiovasc Med 2(2):61, 2001.

Printable version of this article

E-mail this article