Search

A consistently effective transvenous defibrillation system for use in automatic defibrillators could significantly alter the approach to patients at risk of sudden death. Transvenous defibrillation systems that use a right ventricular (RV) electrode only or an RV electrode in combination with a chest patch are relatively inefficient at applying current to the posterolateral left ventricle. An RV electrode combined with a coronary sinus (CS) electrode, however, may improve current distribution to the posterolateral left ventricle. The purpose of this investigation, therefore, was to evaluate the effectiveness and safety of a specially designed transvenous lead system with a CS electrode capable of current delivery to this relatively inaccessible region of the heart. In 20 survivors of cardiac arrest, we determined defibrillation efficacy immediately before defibrillator surgery for monophasic pulses delivered between an RV catheter electrode and a CS catheter electrode system and compared these findings with an RV catheter electrode-thoracic patch defibrillation system. Subsequently, we referenced the efficacy of both transvenous systems to an epicardial patch electrode system at the time of defibrillator implantation. The mean delivered-energy defibrillation threshold for the CS-RV electrode system was 17.5 +/- 7.9 J, which was substantially lower than the RV electrode-thoracic patch system (25.6 +/- 11.4 J, p = 0.0016 [46% more]). Defibrillation threshold voltage was 529 +/- 123 V for the CS-RV electrode system and 647 +/- 164 V (22% more) for the RV electrode-thoracic patch system (p = 0.0013).(ABSTRACT TRUNCATED AT 250 WORDS)

We report a case of mitral valve aneurysm in a 30-year-old man presenting with fatigue, malaise, and fever 10 weeks after emergent aortic valve replacement for endocarditis. The transthoracic echocardiogram demonstrated perivalvular aortic regurgitation, but no abscess cavity was defined. Cardiac magnetic resonance imaging (MRI) revealed a mitral valve leaflet aneurysm. Both aortic and mitral valves were replaced with mechanical prostheses. Pathology of the excised mitral valve showed a focally hemorrhagic aneurysm of the anterior leaflet with myxoid degeneration and focal calcification. Early diagnosis and intervention are important to treat this rare, potentially fatal complication of aortic valve endocarditis. Cardiac MRI provided an accurate and useful preoperative diagnostic evaluation.

This ongoing, longitudinal, multi-center, North American study was designed to evaluate the safety and effectiveness of the On-X valve.The On-X valve was implanted in isolated aortic (AVR) and mitral (MVR) valve replacement patients at nine North American centers. Follow up was 98.6% complete. Anticoagulation compliance was evaluated by collection of international normalized ratio (INR) results in all patients throughout their postoperative follow up. Adverse events were recorded according to the AATS/STS guidance criteria. Hematologic studies were conducted postoperatively to evaluate hemodynamics and hemolysis.In total, 142 AVR and 142 MVR implants were performed; the mean follow up was 4.5 years; total follow up was 1,273 patient-years (pt-yr). At implant, the mean patient age was 59.2 years (range: 28 to 85 years); 71.8% of patients who underwent AVR and 33.1% who underwent MVR were males. Preoperatively, 89.4% of AVR patients and 56.3% of MVR patients were in sinus rhythm. The cardiac disease etiology was primarily stenotic, calcific degeneration in AVR and rheumatic or degenerative regurgitation in MVR. Hemolysis represented by postoperative elevation of serum lactate dehydrogenase was very low (median 217 IU after AVR and 251 IU after MVR at one year (82% AVR and 98% MVR of upper normal). Late adverse event rates were low, most notably thromboembolism (0.9%/pt-yr after AVR; 1.6%/pt-yr after MVR) and thrombosis. Kaplan-Meier event-free rates at five years were correspondingly high. Anticoagulation compliance analysis showed only about 40% of INR readings to be within target ranges postoperatively; thus, the control range achieved was much greater than the desired target, as might generally be expected for clinic-controlled INR.The On-X valve performed well in this study, confirming the original design intent of minimal hemolysis and low adverse event rates.

Background— The first Björk-Shiley convexoconcave (BSCC) prosthetic heart valves were implanted in 1978. The 25th anniversary provided a stimulus to summarize the research data relevant to BSCC valve fracture, patient management, and current clinical options. Methods and Results— Published and unpublished data on the risks of BSCC valve fracture and replacement were compiled, and strategies for identifying candidates for prophylactic valve reoperation were summarized. By December 2003, outlet strut fractures (OSFs), often with fatal outcomes, had been reported in 633 BSCC valves (0.7% of 86 000 valves implanted). Fractures still continue to occur, but average rates of OSFs in 60° valves are now Conclusions— Twenty-five years after the initial BSCC valve implants, fractures continue to occur. Continued monitoring of BSCC patients is needed to track and quantify risks and enable periodic updating of guidelines for patients and their physicians.

Massive organized pericardial hematoma secondary to blunt trauma is extremely rare. We report a case of a 69-year-old man referred with refractory ascites who was found to have a 10 x 6 x 8 cm organized pericardial hematoma with hemodynamic characteristics consistent with constrictive pericarditis. The cause was attributed to blunt chest trauma associated with a fall 3 years previously. Excision of the mass and removal of the constricting epicardium relieved his symptomatology. A review of the literature is included.

OBJECTIVE: To describe the serum concentrations of continuous infusion of cefuroxime for postsurgical prophylaxis of sternal wound infection in patients undergoing coronary artery bypass graft (CABG), and to assess the incidence of sternal wound infection in this population. METHODS: This was a prospective, noncomparative trial involving 54 patients undergoing elective CABG surgery. All patients enrolled in the study received cefuroxime 1.5 g as a single intravenous dose 30 minutes preoperatively, followed by a continuous infusion of 3 g every 24 hours until removal of all central venous catheters. RESULTS: Of the 53 evaluable patients, the mean steady-state cefuroxime serum concentration was 21.6 ± 14.2 μg/mL (range 6.56–59.5). No patient developed a sternal wound infection. The mean treatment duration was 2.58 ± 2.13 days (range 1–13). The median hospital and intensive care unit lengths of stay were six days and 46 hours, respectively. The average antibiotic cost per day was $32.76. CONCLUSIONS: These preliminary results of continuous infusion of cefuroxime 3 g/d for prophylaxis of sternal wound infections in CABG patients indicate that serum concentrations are highly variable, but reliably above the minimum inhibitory concentration for the common anticipated pathogens in this setting. Further comparative trials in a larger number of patients are necessary before this mode of administration can be routinely advocated for prophylaxis.

The effect of electrode polarity on defibrillation thresholds in humans is unknown. This prospective, randomized evaluation of electrode polarity on defibrillation thresholds was performed in 21 survivors of ventricular fibrillation (VF) undergoing cardiac surgery. Defibrillation was always performed with 2 identical large rectangular, wire mesh electrodes positioned over the anterior wall of the right ventricle and the posterolateral wall of the left ventricle. The initial electrode polarity for the left ventricular (LV) electrode was chosen randomly for determination of the defibrillation threshold. Subsequently, electrode polarity was reversed. The defibrillation threshold was defined as the lowest pulse amplitude that would effectively terminate VF with a single discharge delivered 10 seconds after initiation of an episode of VF with alternating current. For each defibrillation pulse, voltage, current, resistance and delivered energy were recorded. Of the 21 patients, 15 (71%) had a lower defibrillation threshold when the LV electrode was positive, 2 patients (10%) had a lower defibrillation threshold when the LV electrode was negative and 4 patients (19%) had equal defibrillation thresholds (within 0.5 J) regardless of polarity. The mean leading edge defibrillation threshold voltage was 370 ± 88 volts when the LV electrode was negative and 320 ± 109 volts (14% less) when the LV electrode was positive (p = 0.014). Mean leading edge defibrillation threshold current was 9.3 ± 3.1 amps when the LV electrode was negative compared to 7.7 ± 3.1 amps (17% less) when the LV electrode was positive (p = 0.0033). There were no differences in resistance with the 2 configurations. The delivered energy defibrillation threshold was 10.1 ± 4.5 J when the LV electrode was negative and 7.8 ± 4.8 J (23% less) when the LV electrode was positive (p = 0.016). Thus, in most patients, the defibrillation threshold is lower when the LV electrode is positive.

Sixteen survivors of cardiac arrest underwent intraoperative comparison of the effectiveness of sequential-pulse and single-pulse defibrillation. Defibrillation was tested alternately with the single-pulse or sequential-pulse technique 10 seconds into an episode of ventricular fibrillation that was induced with alternating current. The sequential-pulse defibrillation technique using truncated exponential pulses was performed with a right ventricular endocardial catheter and a left ventricular epicardial patch electrode. The first pulse was delivered between the right ventricular apical and the superior vena caval electrode on the right ventricular endocardial catheter. The second pulse was delivered between the right ventricular apical electrode and the left ventricular patch electrode 0.2 ms after termination of the first pulse. Single-pulse defibrillation was performed with a standard intracardiac defibrillation system in which a single truncated exponential pulse was delivered across 2 epicardial patch electrodes positioned over the anterolateral right ventricle and the posterolateral left ventricle. During defibrillation threshold determination, voltage and current waveforms were recorded and integrated to determine delivered energy. Average defibrillation threshold leading-edge voltage for the sequential pulse technique was 496 +/- 140 V, compared with 365 +/- 157 V for the single-pulse technique (p less than 0.005). Defibrillation threshold leading-edge current for the sequential-pulse technique was 6.0 +/- 2.3 A, compared with 10.6 +/- 5.1 A for the single-pulse method (p less than 0.0005).(ABSTRACT TRUNCATED AT 250 WORDS)

Nineteen patients with posterior accessory pathways and disabling, refractory arrhythmias, underwent catheter ablation using standard defibrillator pulses at energy settings of 150 to 400 J. Accessory pathway ablation was successful in 13 of 19 (68%). Effective catheter ablation correlated with local ventriculoatrial (VA) intervals determined from the coronary sinus catheter at the site of earliest retrograde atrial activation during orthodromic reciprocating tachycardia. In 12 of the 13 successfully ablated patients, the local VA interval was less than 80 ms. In 4 of the 6 unsuccessfully treated patients, the local VA interval was greater than or equal to 80 ms, p less than 0.01. Transient abnormalities noted with the procedure included sinus bradycardia (3 patients), atrioventricular block (5), accelerated junctional rhythm (3), ectopic atrial tachycardia (2), myocardial depression (1), "ischemic" appearing T-wave inversions (10) and hemodynamically insignificant small pericardial effusions (5) Creatine kinase-MB increased from 3 +/- 2 U/liter to 26 +/- 18 U/liter (p less than 0.001), 4 to 8 hours after ablation. In addition, electrical shorts occurring during the ablation procedure in 2 patients were identified and corrected only with oscilloscopic monitoring of voltage and current waveforms. Significant adverse sequelae were seen in 4 patients. Three patients required sternotomy for control of cardiac tamponade secondary to a ruptured coronary sinus and 1 patient had a small posterior left ventricular infarction related to spasm of a right coronary artery extension branch. Coronary sinus rupture correlated with the ratio of catheter diameter to coronary sinus diameter.(ABSTRACT TRUNCATED AT 250 WORDS)

The risk of coronary arteriography is considerably increased in patients who have left main coronary artery stenosis (LMCAS). Among 1,060 patients undergoing coronary arteriography over a three-year period, 83 were found to have LMCAS and three of these patients died (3.6 percent) during or shortly after the arteriographic procedure. Bypass surgery, however, can now be carried out at a very low risk irrespective of the degree of coronary disease present--there was no hospital mortality in 74 patients with LMCAS undergoing bypass surgery at our institution during this period. In addition to careful attention to detail, techniques that can minimize the risk of both of these procedures in patients with LMCAS include invasive hemodynamic monitoring and various pharmacologic manipulations to prevent myocardial ischemia. In a small percentage of patients, rest angina may prove refractory to inhospital medical treatment, and intraaortic balloon pumping may be necessary to stabilize the conditions of these patients before proceeding with arteriography and myocardial revascularization.

Seventy patients having aortocoronary vein bypass grafting surgery for angina pectoris underwent preoperative invasive exercise testing to symptom limits and again 6 to 14 months postoperatively. Cardiac output was measured using the direct Fick principle. Postoperatively at maximal exercise, there was a 3.11 liters/min (p less than 0.0001) increase in cardiac output in men (n = 61) and a 2.04 liters/min (p less than 0.01) increase in women (n = 9). Patients with complete revascularization showed a significantly greater improvement in cardiac output postoperatively than did those with incomplete revascularization (26 versus 6%, p less than 0.0001). The major reason for the increased maximal cardiac output was a marked increase in heart rate while stroke volume was maintained at the same preoperative level. These findings were true irrespective of preoperative use of beta-adrenergic blocking drugs.

Summary This study has presented the spectrum of postoperative gastrointestinal system complications after open heart surgery at the University of Washington from 1980 through 1983. The frequent necessity for operative intervention and a mortality rate of 17 percent in our study of gastrointestinal complications in patients who have undergone open heart surgery indicates the need for early diagnosis and treatment. The data suggest that bypass times approaching 100 minutes and the presence of postoperative cardiogenic shock are important risk factors in the development of such complications in elective cardiac surgery patients. An incidence of gastrointestinal complications of 8.6 percent in those undergoing repair of acute aortic dissections makes gastrointestinal complaints particularly suspicious in this subgroup.

To assess the efficacy of surgical revascularization for postinfarction angina within 30 days of acute infarction, the clinical course of 103 patients treated surgically from January 1979 to July 1982 was reviewed. There were 84 men (82%) and 19 women (18%) with a mean age of 58 years (range 34 to 80). Group A (11 patients) underwent surgery within 24 hours of infarction, Group B (21 patients) within 7 days and Group C (71 patients) within 30 days. Eighty-four patients (82%) had subendocardial infarctions and 19 patients (18%) had transmural infarction. Transmural infarction was more common in patients in Group A (36%) than in those in either Group B (19%) or Group C (15%). There were two deaths, both in Group C (1.9%), within 30 days of surgery. The use of intraaortic balloon or inotropic support and the occurrence of major arrhythmias or perioperative infarction was noted in 30 patients (29%)(64% in Group A, 33% in Group B and 18% in Group C). The average time in the intensive care unit was 3.2 days, with an average total hospital stay after surgery of 8.3 days. Late follow-up (mean 15.4 months, range 1 to 39) is complete for 97 patients (97%). There were no late myocardial infarctions and 93 patients (96%) were essentially free of angina. The only late death (1.0%) was caused by septicemia from delayed sternal wound infection. This study suggests that myocardial revascularization within the first 30 days after myocardial infarction can be accomplished with an acceptable operative mortality in selected patients with postinfarction angina refractory to medical management.

The effect of initially ineffective defibrillation pulses on subsequent defibrillation success is not known. Therefore, the voltage, current and energy at the defibrillation threshold were compared with the defibrillation rescue pulse voltage, current and energy that terminated ventricular fibrillation when an ineffective pulse just below the defibrillation threshold had been used initially. This lower amplitude ineffective pulse was termed a “subdefibrillation threshold” pulse. The pulse that restored sinus rhythm after the subdefibrillation threshold pulse was delivered was termed the “subdefibrillation threshold rescue pulse.” This comparison was undertaken, intraoperatively, in 14 out-of-hospital cardiac arrest survivors using a sequential-pulse catheter-patch defibrillation system. Each of the 14 patients required higher voltage, current, delivered energy and stored energy for defibrillation with the subdefibrillation threshold rescue pulse than with the defibrillation threshold pulse. The defibrillation threshold voltage was 451 ± 127 volts compared with a subdefibrillation threshold rescue voltage of 585 ± 147 volts (p < 0.00002). The defibrillation threshold current was 5.5 ± 2.4 amps compared with a subdefibrillation threshold rescue current of 7.2 ± 2.7 amps (p < 0.00001). Delivered and stored energies were 10.9 ± 7.4 and 12.3 ± 7.2 J, respectively, for the defibrillation threshold pulse and were 17.6 ± 9.4 J (p < 0.00002) and 20.5 ± 9.3 J (p < 0.00005), respectively, for the sub-defibrillation threshold rescue pulse. It is concluded that the risk of requiring considerably higher energies than anticipated for defibrillation must be incurred, should defibrillation fail because of an initially insufficient defibrillation pulse. Such an occurrence might result in death should the energies necessary for defibrillation exceed the output limits of the implantable device.

Sixteen out-of-hospital survivors of ventricular fibrillation underwent a prospective, randomized, intraoperative comparison of sequential pulse and single pulse defibrillation with use of two distinct electrode systems and waveform shapes currently available for clinical use. Defibrillation was tested alternately with either the single pulse or the sequential pulse system 10 s into an episode of ventricular fibrillation. Sequential pulse defibrillation was performed with two 4 ms truncated exponential pulses of constant duration delivered to three equally spaced oval epicardial patch electrodes composed of concentric coils. The posterior left ventricular electrode served as the common cathode. The first anode was over the anterior right ventricle and the second anode was over the anterior left ventricle. Single pulse defibrillation was performed with the standard intracardiac defibrillation system with use of a single truncated exponential pulse with a fixed 65% tilt delivered across two rectangular, wire mesh epicardial patch electrodes positioned over the anterior right ventricle and posterolateral left ventricle. During defibrillation threshold determination, voltage and current waveforms were recorded and used to determine pulsing resistance and delivered and stored energy. Average defibrillation threshold leading edge voltage for the single pulse technique was 273 ± 101 V compared with 246 ± 67 V (11% less) for the sequential pulse technique (p = 0.136). Defibrillation threshold leading edge current for the single pulse technique was 6.7 ± 2.5 A compared with 5.2 ± 1.7 A (29% less) for the sequential pulse method (p = 0.005). The defibrillation threshold delivered energy was 5.6 ± 4.0 J for the single pulse technique and 3.5 ± 1.8 J (38% less) for the sequential pulse technique (p = 0.021). The defibrillation threshold stored energy for the single pulse technique was 6.3 ± 4.3 J compared with 4.0 ± 2.0 J (37% less) for the sequential pulse technique (p = 0.013). In survivors of cardiac arrest, the clinically available sequential pulse three electrode defibrillation system is at least as viable as the standard single pulse two electrode system for defibrillation of the heart. In some individuals, the sequential pulse system utilizes substantially lower current and energy to defibrillate than does the single pulse system.

Twenty-three patients with bacterial endocarditis and mycotic aneurysms of the aortic annulus were treated between 1978 and 1985. There were 18 men and 5 women ranging from 24 to 72 years old. All patients had congestive heart failure and positive blood cultures as a complication of the endocarditis and were in New York Heart Association (NYHA) Functional Class III or IV. The aneurysm complicated late prosthetic valve endocarditis in 7 patients and native valve endocarditis in 16. The most common infecting organisms were streptococci (12 patients) and staphylococci (7 patients). The noncoronary sinus was the most frequent site for aneurysm formation. Following debridement of the abscess cavity, the orifice of the aneurysm was closed with a patch of Dacron in 20 patients and autologous pericardium in 3. A prosthetic valve (18 bioprosthetic and 5 mechanical) was secured to the noninfected portion of the native annulus and to the patch at the level of annulus. There were 3 deaths, 1 perioperative and 2 late, each without evidence of residual infection or aortic insufficiency. There are 20 late survivors (87%). After a mean follow-up of 1 year, all patients are in NYHA Functional Class I. Patch closure of mycotic aneurysms involving the aortic annulus permits aggressive debridement of the abscess cavity and affords closure of the orifice without tension. The prosthetic valve can be seated at the level of the native annulus, thus avoiding complicated reconstructive procedures of the aortic root and coronary arteries. This technique is an effective alternative in selected cases of mycotic aneurysms involving the aortic annulus.

During a 30-month period, 51 patients underwent mitral valve replacement. There were 3 hospital deaths (5.9%), 2 of which were due to ventricular rupture. The 3 patients who died were among 13 patients in whom mitral valve replacement was combined with tricuspid or aortic valve operation or both. Postmortem findings in the 2 patients who died of ventricular rupture showed that the ventricular tears were located between the atrioventricular groove and the unresected papillary muscle stumps, in an area of ventricle formerly tethered by the posterior chordae tendineae. In the last 14 patients in the series, the posterior leaflet of the mitral valve and its chordae tendineae were left intact, and there was no mortality or prosthetic valve dysfunction. In patients with myxomatous or ischemic disease, the posterior leaflet was left completely intact. For patients with fibrocalcific rheumatic disease, we have developed a technique of partial excision and debridement of the posterior leaflet, preserving the intermediate and basal chordae tendineae attachments. With the techniques described, preservation of all or part of the posterior leaflet and its chordae tendineae does not appear to interfere with prosthetic valve function and, by reducing the risk of ventricular rupture, should enhance survival after mitral valve replacement.

The effect of ventricular fibrillation duration in humans on defibrillation efficacy as it pertains to the time of intervention of an automatic implantable defibrillator is unknown. If a difference in defibrillation efficacy exists in the early period after ventricular fibrillation onset, it may affect algorithms used by antiarrhythmic devices for arrhythmia detection and therapy. Therefore, a prospective, randomized evaluation was performed of the effect of ventricular fibrillation durations of 10 s and 20 s on defibrillation thresholds in 10 survivors of sudden cardiac arrest undergoing implantation of an automatic cardioverter defibrillator. The initial duration of ventricular fibrillation was chosen randomly. Subsequently, each patient served as his or her own control for the alternate duration of ventricular fibrillation to that chosen initially. The mean leading edge defibrillation threshold voltage was 411 ± 114 V when ventricular fibrillation persisted for 10 s and 419 ± 125 V when it persisted for 20 s (p = 0.73). The mean defibrillation threshold current was 11.4 ± 2.8 A when ventricular fibrillation persisted for 10 s and 11.4 ± 3.2 A when it persisted for 20 s (p = 0.97). The delivered energy defibrillation threshold was 11.5 ± 5.9 J when ventricular fibrillation persisted for 10 s and 12.0 ± 6.9 J when it persisted for 20 s (p = 0.67). These findings show that the defibrillation threshold does not change between 10 and 20 s of ventricular fibrillation in out-of-hospital survivors of cardiac arrest at the time of surgical implantation of an automatic defibrillator. The data may have influence on the programming of defibrillator detection algorithms.

Biphasic waveforms have been suggested as a superior waveform for ventricular defibrillation. To test this premise, a prospective randomized intraoperative evaluation of defibrillation efficacy of monophasic and biphasic waveform pulses was performed in 22 survivors of out of hospital ventricular fibrillation who were undergoing cardiac surgery for implantation of an automatic defibrillator.The initial waveform used in a patient for defibrillation testing, either monophasic or biphasic, was randomly selected. Subsequently, each patient served as his or her own control for defibrillation testing of the other waveform. The defibrillation threshold was defined as the lowest pulse amplitude that would effectively terminate ventricular fibrillation with a single discharge delivered 10 s after initiation of an episode of ventricular fibrillation induced with alternating current. Each defibrillation pulse was recorded oscilloscopically, and defibrillation pulse voltage, current, resistance and stored energy were measured.Fifteen (68%) of the 22 patients had a lower defibrillation threshold with the biphasic pulse, 3 (14%) had a lower threshold with the monophasic pulse and 4 (18%) had equal defibrillation thresholds (within 1.0 J) regardless of waveform. The mean leading edge defibrillation threshold voltage was 317 ± 105 V when the monophasic pulse was used and 267 ± 102 V (16% less) when the biphasic pulse was used (p = 0.008). Mean leading edge defibrillation threshold current was 7.9 ± 3.7 A when the monophasic pulse was used and 6.8 ± 3.8 A (14% less) when the biphasic pulse was used (p = 0.051). There were no differences in system resistance with the two waveforms: the leading edge resistance for the monophasic pulse was 45.3 ± 17.5 and 44.1 ± 14.4 ω for the biphasic pulse (p = 0.183). The stored energy defibrillation threshold was 8.5 ± 6.1 J when the monophasic pulse was used and 6.3 ± 5.2 J (26% less) when the biphasic pulse was used (p = 0.028). These findings show that in the majority of patients, biphasic pulse defibrillation is more efficient than monophasic pulse defibrillation.

In a series of 15 studies in dogs, sternocleidomastoid muscle was used to replace deficits created in left ventricular myocardium and sternohyoid muscle was used to replace portions of right myocardial wall. The five right ventricular autotransplants resulted in a 100% surgical success rate, with animals electively killed between 3 and 55 weeks after surgery. In 10 left ventricular studies excision of areas varying from 12 X 46 mm to 30 X 60 mm and incisions of from 40 mm to 70 mm in length were performed. Left ventricular studies resulted in a 60% surgical success rate, with clinically healthy animals being killed for study between 2 weeks and 50 weeks after surgery. Animals surviving the critical surgical recovery period showed no loss of weight or changes in activity. Gross findings at autopsy confirmed the viability of the skeletal muscle transplants. Borders were well healed and the grafted tissue was pliable. Histologic studies suggest that revascularization of skeletal muscle occurred from the myocardial side, and that there were healthy myocardial and skeletal muscle fibers at the junction, with evidence of regeneration.

Some of the reported problems with the use of the automatic implantable defibrillator (AID) include inappropriate discharge, 1–3 failure to defibrillate because of excessively high defibrillation thresholds 3–5 and failure to sense ventricular fibrillation (VF) in patients who also have pacemakers. 1,2 In this report, we describe another potential problem: the failure of an AID to sense VF because of variable bipolar epicardial electrogram amplitudes during VF.