Author + information
- Received March 18, 1996
- Revision received July 8, 1997
- Accepted July 16, 1997
- Published online November 1, 1997.
- Ngai-Sang Lok, MBA,
- Chu-Pak Lau, MD, FRCP, FACCA,*,
- Hung-Fat Tse, MBBS, MRCPA and
- Gregory M Ayers, MD, PhDB
- ↵*Professor Chu-Pak Lau, Department of Medicine, University of Hong Kong, Queen Mary Hospital, Hong Kong, People’s Republic of China.
Objectives. The objectives of this study were 1) to evaluate the effect of different right atrial electrode locations on the efficacy of low energy transvenous defibrillation with an implantable lead system; and 2) to qualitate and quantify the discomfort from atrial defibrillation shocks delivered by a clinically relevant method.
Background. Biatrial shocks result in the lowest thresholds for transvenous atrial defibrillation, but the optimal right atrial and coronary sinus electrode locations for defibrillation efficacy in humans have not been defined.
Methods. Twenty-eight patients (17 men, 11 women) with chronic atrial fibrillation (AF) (lasting ≥1 month) were studied. Transvenous atrial defibrillation was performed by delivering R wave-synchronized biphasic shocks with incremental shock levels (from 180 to 400 V in steps of 40 V). Different electrode location combinations were used and tested randomly: the anterolateral, inferomedial right atrium or high right atrial appendage to the distal coronary sinus. Defibrillation thresholds were defined in duplicate by using the step-up protocol. Pain perception of shock delivery was assessed by using a purpose-designed questionnaire; sedation was given when the shock level was unacceptable (tolerability threshold).
Results. Sinus rhythm was restored in 26 of 28 patients by using at least one of the right atrial electrode locations tested. The conversion rate with the anterolateral right atrial location (21 [81%] of 26) was higher than that with the inferomedial right atrial location (8 [50%] of 16, p < 0.05) but similar to that with the high right atrial appendage location (16 [89%] of 18, p > 0.05). The mean defibrillation thresholds for the high right atrial appendage, anterolateral right atrium and inferomedial right atrium were all significantly different with respect to energy (3.9 ± 1.8 J vs. 4.6 ± 1.8 J vs. 6.0 ± 1.7 J, respectively, p < 0.05) and voltage (317 ± 77 V vs. 348 ± 70 V vs. 396 ± 66 V, respectively, p < 0.05). Patients tolerated a mean of 3.4 ± 2 shocks with a tolerability threshold of 255 ± 60 V, 2.5 ± 1.3 J.
Conclusions. Low energy transvenous defibrillation with an implantable defibrillation lead system is an effective treatment for AF. Most patients can tolerate two to three shocks, and, when the starting shock level (180 V) is close to the defibrillation threshold, they can tolerate on average a shock level of 260 V without sedation. Electrodes should be positioned in the distal coronary sinus and in the high right atrial appendage to achieve the lowest defibrillation threshold, although other locations may be suitable for certain patients.
The efficacy and safety of converting atrial fibrillation (AF) by delivering low energy shocks between intracardiac catheters with shock vectors encompassing both the left and right atria have been demonstrated in animal and human studies [1–10]. These studies suggest that the use of intracardiac catheter-based electrodes in combination with an implantable device would be feasible, but patient and physician acceptance of such a device is yet to be determined. Previous studies [11–13]have shown the relation between the defibrillation threshold with different defibrillation waveforms, size of defibrillation electrodes and electrode configurations. A lower defibrillation threshold may have a significant impact on device size and longevity and, possibly, on patient tolerance. To date, the most efficient and most commonly used method of atrial defibrillation in humans has been delivery of biphasic shocks between electrodes positioned in the coronary sinus and the lateral right atrium.
Unlike patients with ventricular fibrillation and an implantable ventricular defibrillator, who may become hemodynamically unstable during the arrhythmia, patients with paroxysmal AF are likely to be conscious during their arrhythmia. Thus, the ability of patients to tolerate defibrillation shocks without sedation remains one of the main unanswered questions concerning use of an implantable atrial defibrillator. In a study using a 20-V step-up incremental shock protocol for transvenous defibrillation beginning at a shock level of 20 V, the mean tolerated energy level of a shock delivered without sedation was ∼120 V, a level far below the mean successful voltage (237 V) for restoration of sinus rhythm . This finding raised concern that sedation will be routinely required for successful defibrillation. However, the reported method of threshold determination, although potentially more accurate than using a higher starting energy level, may not be applicable to device-mediated patient treatment. Clinically, an initial shock level that is moderately high—that is, closer to the level likely to result in successful arrhythmia conversion—should decrease the number of shocks delivered and thus might make this form of therapy more acceptable.
We hypothesized that defibrillation thresholds would differ significantly with different atrial electrode locations when an implantable lead system was used clinically. Our secondary hypothesis was that patients would tolerate shocks of a higher intensity than values previously reported if the starting shock intensity was closer in intensity to the predicted defibrillation threshold.
1.1 Patient Selection (Table 1)
All patients gave written informed consent before the procedure, and the protocol was approved by the ethics committee of the University of Hong Kong. Transvenous atrial defibrillation was performed in 28 patients (17 men, 11 women) with electrocardiographically documented AF for ≥1 month (range 1 to 192 months, exact duration unknown in 2 patients). Patients with the following conditions were excluded from study: 1) reversible causes of AF, 2) significant valvular heart disease, 3) unstable angina or recent onset myocardial infarction, 4) class III or IV heart failure, and 5) evidence of left atrial clot or thrombi. The patients had a mean age of 62 ± 9 years and a mean left atrial diameter of 4.5 ± 0.6 cm. A detailed clinical examination, chest radiography and transthoracic and transesophageal echocardiography were routinely performed. Other clinical characteristics of the patients and the electrode locations tested are summarized in Table 1. All antiarrhythmic agents were stopped at least 5 half-lives before the procedure.
All patients were treated with oral anticoagulation using warfarin to achieve an international normalized ratio (INR) of 2 to 3 for ≥3 weeks before the procedure. Oral anticoagulation was discontinued 2 days before the procedure and was replaced with an intravenous heparin infusion. The heparin infusion was stopped 4 h before and restarted 4 h after the procedure. The INR was checked daily and immediately before the procedure; the minimal INR value deemed safe for venous puncture was <1.5. Oral anticoagulation was restarted after the cardioversion and was continued for ≥3 months. Heparin infusion was discontinued when the therapeutic INR of 2 to 3 was again achieved.
1.3 Defibrillation Lead Placement
Defibrillation testing was performed with the patient in a fasting state without sedative or hypnotic premedication. Local anesthesia with bupivacaine (0.25%) was given before venous puncture. Two 8F sheaths were introduced into the left subclavian vein from separate punctures. Implantable defibrillator leads were positioned in the distal coronary sinus and right atrium; exact locations for testing were determined on the basis of a randomized location order.
The electrode placed in the coronary sinus was on a transvenous passive fixation lead (model 7107, Perimeter CS lead, InControl, Inc.), with a nominal defibrillator coil length of 6 cm and an electrode surface area of 4.7 cm2. The defibrillation coil has a pigtail-like spring coil of 2.5 turns that was deployed when the stylet was withdrawn (Fig. 1). With use of a stiff stylet shaped with a gentle curve, the lead was advanced as distally as possible into the coronary sinus, making certain that it had not advanced too anterior to beyond the margin of the left atrium and was therefore considered advanced too far and into the descending vein onto the ventricle. The philosophy behind the positioning of this lead was to mimic the clinical implantation of such a lead and to use this same coronary sinus lead location to evaluate each of the right atrial lead locations.
The second defibrillation lead (model 7203, Perimeter RA lead, InControl Inc.) for the right atrium was an active fixation screw-in lead, with a nominal defibrillation coil length of 6 cm and an electrode surface area of 5.2 cm2. The right atrial defibrillation lead was advanced under fluoroscopic control while maintaining counterclockwise rotation of the lead body. Once the lead tip reached the right atrium, the straight stylet was withdrawn. With the use of a manually shaped stylet (either J-shaped for the appendage location or with a gentler curve for the other locations), the tip of the lead was positioned in one of the three right atrial electrode locations for defibrillation threshold testing.
The three electrode locations were determined by dividing the atrial chamber into three approximately equal but discrete parts based on the fluoroscopic silhouette of the right atrium. These three electrode locations were defined as the high right atrial appendage, anterolateral right atrium and inferomedial right atrium (Fig. 2). The lead was advanced so that the coil electrode was located within the atrium (upper third for the high right atrial appendage location, lateral third for the anterolateral right atrial location or the inferior third of the right atrium, and in contact with the inferior right atrial wall, for the inferomedial right atrial location). After the desired right atrial location was reached, the lead tip was affixed to the right atrial wall by turning the lead body clockwise over the stylet with three to five complete turns. The stylet was subsequently removed and the coil electrode adjusted with lead advancement and withdrawal so that the coil electrode was located within the desired third of the right atrium. As with the coronary sinus lead, this study was designed to mimic the clinical use of the lead system; therefore, electrode location was made as consistent as possible within the clinical confines of the anatomy of a given patient to allow clinical comparison of three electrode locations.
Two 6F pacing catheters were also positioned. One was placed in the right atrium (for AF induction) and one in the right ventricle (for shock synchronization); both were advanced from the right femoral vein.
1.4 External Atrial Defibrillator
The defibrillation leads were connected to a custom-built external defibrillator (XAD, InControl, Inc.), whereas the temporary catheters were attached to a programmable stimulator (model 5328, Medtronic Inc.) and to the recording equipment. Surface electrocardiogram, atrial and ventricular electrograms and arterial blood pressure were continuously monitored by using a chart recorder (Mingograf 7, Siemens Ltd, Sweden) at a paper speed of 50 mm/s. A test shock of 20 V was first delivered to assess R wave synchronization and the integrity of the defibrillation system.
R wave-synchronized, 3/3 ms duration biphasic shocks were then delivered, starting with a shock intensity of 180 V. The shock intensity was increased in steps of 40 V until sinus rhythm was restored (atrial defibrillation threshold) or until a shock of the maximal intensity deliverable from the device (400 V) was reached. The shock with the lowest intensity that resulted in successful conversion was considered the defibrillation threshold. To allow for statistical analysis of thresholds, one step (40 V) higher than the maximal intensity applied was used to represent the defibrillation threshold when shocks failed to convert AF with use of a particular right atrial lead location. If sinus rhythm was restored and sustained for ≥1 min, AF was reinduced by rapid atrial pacing. The defibrillation threshold was then redetermined by using the same right atrial lead location under test, again using the 40-V step-up protocol; however, this second threshold determination used a starting intensity that was two steps (80 V) lower than the previously determined threshold. Should any of the initial shocks (either 180 or 80 V less than the first threshold) successfully convert the AF, the shock intensity was decreased by two steps (80 V) and the 40-V step-up protocol was started from this initial failed shock level. After determination of two thresholds for successful defibrillation or after the delivery of two 400-V shocks that failed to convert AF, the right atrial defibrillation lead was repositioned to the next location and the defibrillation thresholds were determined by using the same method.
1.5 Lead Sequence Testing
The right atrial electrode was randomly positioned in both the anterolateral and inferomedial right atrial locations in 9 patients, in both the anterolateral right atrial and high right atrial appendage locations in 10 patients and in all three locations in 9 patients. In some patients, because of the length of the procedure, only two lead locations were tested.
1.6 Pain Perception Assessment
The method of assessing discomfort level and the patient discomfort questionnaire were explained to all patients in detail before the defibrillation procedure. Shock perception was assessed by using a 10-point visual analog scale. On this scale, 0 represented “not felt,” whereas 10 represented “extremely uncomfortable.” The shock intensity that resulted in a perception score of 10 was considered the tolerability threshold. All patients were warned before each shock application, but the shock intensity was not disclosed. Once the tolerability threshold was reached, patients were asked to complete a discomfort questionnaire (Table 2). Patients were then sedated with midazolam (0.05 mg/kg body weight) and pethidine (0.5 mg/kg); additional doses were given as required.
1.7 Statistical Analysis
Repeated measures analysis of variance (ANOVA) was used to compare the defibrillation threshold for the different right atrial electrode locations and to compare the first and second thresholds determined for each of these locations. Rates of successful conversion for the different right atrial electrode locations were compared by using the Fisher exact test. All results are expressed as mean value ± 1 SD; a p value < 0.05 was considered statistically significant.
2.1 Defibrillation Efficacy and Threshold
Evaluation of overall efficacy of the procedure revealed that sinus rhythm was restored in 26 of the 28 patients. The procedure did not induce ventricular arrhythmia or result in other complications. The number of patients whose arrhythmia was converted by using the inferomedial right atrial electrode location (8 [50%] of 16) was less than the number whose arrhythmia was converted by using either high right atrial appendage or anterolateral right atrial electrode locations (16 [89%] of 18 and 21 [81%] of 26, respectively, p < 0.05). The latter two electrode locations did not differ significantly in the number of patients with arrhythmia conversion (p > 0.05). The mean defibrillation thresholds for the high right atrial appendage, anterolateral right atrial and the inferomedial right atrial electrode locations are shown in Fig. 3; all were significantly different with respect to energy (3.9 ± 1.8 vs. 4.6 ± 1.8 vs. 6.0 ± 1.7 J, respectively, p < 0.05) and voltage (317 ± 77 vs. 348 ± 70 vs. 396 ± 66 V, respectively, p < 0.05). However, individual thresholds for a given patient did not always follow the trend of the means for the different right atrial electrode locations (Table 1). There was no significant difference in when the first and second defibrillation thresholds were compared for the same right atrial electrode location (p > 0.05).
2.2 Pain Perception in Transvenous Atrial Defibrillation
Four patients tolerated the maximal shock level (400 V) without sedation, and one patient requested sedation before positioning of the leads and shock delivery. Sinus rhythm was restored in two patients (at 220 and 400 V, respectively) before the tolerability threshold was reached. Two patients had successful defibrillation at their respective tolerability thresholds of 220 and 340 V. The need for sedation at different shock levels is shown in Table 3. After excluding patients in whom AF was not converted even at 400 V without sedation and patients who had successful defibrillation before the tolerability threshold was reached, the tolerability threshold in the remaining 22 patients was 2.5 ± 1.3 J, 255 ± 60 V (range 1 to 5.6 J, 180 to 380 V). On average, patients tolerated 3.4 ± 2 shocks either before cardioversion was achieved (5 patients) or before the patient requested sedation (17 patients).
All patients in whom the questionnaire was used (n = 19) described the perception at the tolerability threshold as a dull discomfort mainly localized in the anterior chest (Table 2). In 18 of the 19 patients, discomfort disappeared 5 to 30 s after shock delivery; in 1 patient the discomfort lasted ∼1 min. Three patients reported mild to moderate shortness of breath.
3.1 New Findings of This Study
The two new findings in this study relate to the clinical use of an implantable atrial defibrillator that utilizes a right atrium to coronary sinus shock vector. This study demonstrated that the site selected for placement of the right atrial defibrillation electrode has a significant impact on the atrial defibrillation threshold. Therefore, this study can provide guidance in the clinical positioning of such a lead and its electrode. Second, comparison of the shock tolerability data collected in this study with the results of other studies of transvenous atrial defibrillation suggests that the number of shocks may have a greater effect than shock intensity on patient perception of shock tolerability. This finding has important implications for the determination of patient acceptance of internal defibrillation therapy for AF and the clinical use of an implantable atrial defibrillator.
3.2 Defibrillation Threshold and Lead Locations
The energy requirement for transvenous atrial defibrillation is only ∼2% to 5% of the energy required for transthoracic defibrillation [1–8]and ∼2% of the energy requirement in an early investigation using an intracardiac defibrillation catheter and an external plate on the left posterior chest. Shock delivery with different electrode locations has been found to affect the distribution of energy and, therefore, the defibrillation threshold. In animal studies [11, 12], it has been demonstrated that right to left shock vectors encompassing both atria had significantly lower defibrillation thresholds than right-sided vectors, and shock delivery between the right atrium and coronary sinus was more efficacious than shock delivery using other vectors. A recent study assessed the effect of different lead locations on defibrillation threshold in a group of patients with heterogeneous cardiac diseases. The investigators found that a mean defibrillation threshold of 9.9 J was required in three randomized positions tested (right atrium and skin patch, right atrium and right ventricle, and right atrium to superior vena cava, which had the lowest defibrillation threshold). A trend toward a lower defibrillation threshold in the right atrium to left pulmonary artery (five patients) or coronary sinus vector (two patients) was seen in this nonrandomized study, although the defibrillation thresholds were only 7.5 and 2 J in the two patients with the right atrium to coronary sinus configuration. Thus, the available data on lead location in humans suggest that a right atrium to coronary sinus vector may be associated with the lowest defibrillation threshold. However, none of these studies examined the site of the right atrium associated with the lowest defibrillation threshold, which may have a significant impact on patient tolerability and battery longevity. To our knowledge, our study is the first to examine the effects of different right atrial electrode locations. We demonstrated that high right atrial appendage electrode location appears to have, on average, a lower atrial defibrillation threshold than that of an anterolateral right atrial electrode location. The inferomedial right atrial electrode location had the lowest threshold and also the smallest number of patients whose arrhythmia was converted with use of this electrode location. This finding is at odds with the results from a study in sheep in which the superior right atrial location had a higher defibrillation threshold (1.25 ± 0.52 J) than that of a low right atrial appendage or anterolateral right atrial location (0.92 ± 0.48 and 0.77 ± 0.31 J, respectively, p < 0.05). This difference may be due to the difference between sheep and human anatomy or to the use of temporary catheters in the animal study and leads of a permanent type in this clinical study.
Although the high right atrial appendage electrode location in our study resulted in the lowest mean defibrillation threshold, it is important to note that this finding was not consistent for all patients tested. Second, it is necessary to consider, in addition to the threshold data, the number of patients whose arrhythmia could be converted by using each of the three right atrial electrode locations tested. These two issues taken into account, we believe that our study data can serve as a guide for the clinical placement of right atrial defibrillation leads. On the basis of our results, we recommend that clinicians first attempt to locate the right atrial electrode in the high right atrial appendage; if thresholds are inadequate for defibrillation or tolerability with this electrode location, the anterolateral right atrium should then be tried. The inferomedial right atrial electrode location, given the smaller number of patients in this study with successful arrhythmia conversion and a higher mean threshold, would be unlikely to provide either an adequate threshold or an adequate likelihood of successful conversion. Therefore, we anticipate that this electrode location would be used only in rare situations. Lastly, the threshold for an electrode location must be considered against the difficulty in lead placement, the location clinically obtainable in a given patient with the coronary sinus electrode, and stresses on the tip of the right atrial lead. The lateral right atrium is potentially the easiest location to achieve, requiring less stress on the lead tip than other sites; however, this location is less distant from a more proximal coronary sinus electrode location, which included a lesser amount of atrial myocardium in the elevated field.
The atrial defibrillation threshold we obtained was only half the value obtained in a similar clinical study by Saksena et al. , but it is similar to thresholds reported in a study using temporary catheters in the right atrium to coronary sinus vector. This finding may be due to the fact that the majority of patients in our study did not have major underlying cardiac diseases. However, the mean left atrial diameter in our study, 4.5 cm, was identical to that reported by Saksena et al. . Additionally, chronic AF was present in all patients in our study but in only 50% of the patients of Saksena et al. . Preliminary data [6, 10]suggested that chronic AF is associated with a higher defibrillation threshold. Thus, we believe that the lower defibrillation threshold observed in our study may be attributed to the lead locations we used. The study of patients without significant structural heart disease is a logical first step in the investigation of an implantable atrial defibrillator. Further study will be required to assess the efficacy of this technique in different patient groups.
3.3 Tolerability of Low Energy Transvenous Atrial Defibrillation
Previous data on the tolerability of low energy transvenous atrial defibrillation are limited. Murgatroyd et al. performed low energy transvenous defibrillation in 4 patients with spontaneous AF and 18 patients with induced AF and reported a maximal tolerable shock intensity of 180 V (1.0 J). Apart from 2 of their patients who had sinus rhythm restored with “severely uncomfortable” shocks (140 V) and 3 who could not tolerate the smallest shock intensity (10 or 20 V), the other 17 patients required sedation before the threshold voltage (237 ± 55 V) was reached. Another study , using catheters in the right atrial appendage and the coronary sinus for converting AF in 10 patients, also reported a fairly low tolerability. All patients in that study required sedation at a shock energy intensity of <2 J. Tolerability data were also limited in a third study ; only 2 of 18 patients could tolerate the effective shock without sedation. The tolerability in our patients was significantly higher than that of the patients in these studies. In our study, the mean tolerability threshold was not too distant from the defibrillation threshold, and a few patients tolerated the maximal voltage tested (400 V) without sedation.
Because the study of Murgatroyd et al. used a 20-V initial shock intensity, more shocks were required before sinus rhythm was restored. Additionally, in that study, as in our study, two to three shocks were delivered before the tolerability threshold was reached. Therefore, the tolerability of atrial defibrillation shocks may be related more to the cumulative effect of shock delivery—that is, the total number of shocks delivered—than to the intensity of individual shocks. Therefore, it would appear that patient tolerability is improved if the first shock level is close to or slightly above the predicted defibrillation threshold. A gradual step-up protocol with a low starting intensity used for therapeutic conversion in patients undergoing internal atrial defibrillation or with an implantable atrial defibrillator will probably be poorly tolerated. As most patients will tolerate a mean of two to three shocks, a subsequent “rescue” shock should probably be of a sufficiently higher intensity to ensure successful defibrillation. With the shock delivery methodol employed in this study, we demonstrated improved patient tolerability by using this concept—namely, increasing the shock intensity so that the number of shocks can be limited.
Psychologic factors, such as patient preparation in the form of a low intensity warning shock and patient-perceived benefit from shock delivery, may also play a critical role with respect to patient tolerability. Methods that lower the defibrillation threshold, such as concomitant administration of an antiarrhythmic agent or the use of rapidly absorbed medications to relieve shock-induced discomfort, may further enhance patient acceptability of a low energy implantable atrial defibrillator. Additionally, technologic advances in waveform and electrodes may further increase the tolerability of internal atrial defibrillation.
3.4 Limitations of the Study
Our study was concerned with atrial defibrillation threshold in patients with chronic AF, and it is likely that a lower defibrillation threshold will be obtained for acute onset AF, as has been shown by using temporary lead systems . However, the higher energy requirements of chronic AF enabled us to assess patient tolerability to higher intensity shocks, thereby adding important data to the clinical application of this technique. Although the reduced defibrillation energy requirement—3.9 rather than 6.0 J—may still be perceived by the patients, it is conceivable that a similar reduction would contribute to better patient tolerance in patients with acute onset AF. Further studies will be needed to evaluate the long-term changes in defibrillation threshold and tolerability in patients with an implantable atrial defibrillator system. This study was underpowered to assess the effect of lead location on discomfort, a factor that might theoretically have an effect on clinical practice as important as that of defibrillation threshold.
Because our study utilized the initial, chronic episode of AF in determining the first defibrillation threshold, it could be argued that this initial threshold would be higher than the thresholds obtained for subsequent induced episodes of AF. However, previous studies by our group in similar patients have shown that thresholds are not higher for spontaneous, longer-lasting episodes than for induced AF. Additionally, this study was designed so that each threshold was redetermined at each right atrial location site, thereby determining a threshold for both the long-lasting episode and for induced AF. Statistical analysis with ANOVA of the initial threshold versus subsequent thresholds, allowing for the difference in right atrial location, showed that the initial threshold was not significantly different from the subsequent threshold in our patient group (p > 0.05). Therefore, concerns surrounding this issue should not diminish the strength of the conclusions and findings reported herein.
Low energy transvenous defibrillation with an implantable defibrillation lead system is an effective treatment for restoring sinus rhythm in patients with AF. Defibrillation leads should be implanted so that the electrode is located in the high right atrial appendage or, secondarily, in the anterolateral right atrium when used with a second lead in the distal coronary sinus. When a moderate rather than a low initial shock level is used, most patients can tolerate the defibrillation shock level of 260 V and two to three shocks. To improve patient tolerability, a starting shock level close to the defibrillation threshold should be used in a shock delivery protocol.
☆ This study was partially funded by InControl Inc.
- atrial fibrillation
- analysis of variance
- international normalized ratio
- Received March 18, 1996.
- Revision received July 8, 1997.
- Accepted July 16, 1997.
- The American College of Cardiology
- ↵Kumagai K, Yamanouchi Y, Tashiro N, Hiroki T, Arakaw K. Low energy synchronous transcatheter cardioversion of atrial flutter/fibrillation in the dog. J Am Coll Cardiol 1990;16:497–50.
- Powell A,
- Garan H,
- McGovern BA,
- Fallon JT,
- Krishnan SC
- Ayers GM,
- Alferness CA,
- Ilina M,
- et al.
- Lok NS,
- Lau CP,
- Tang YW,
- Ho SW
- Murgatroyd FD,
- Slade AKB,
- Sopher SM,
- Rowland E,
- Ward DE,
- Camm AJ
- Lok NS, Lau CP, Ho DSW, Tang YW. Hemodynamic effects and clinical determinants of defibrillation threshold for low energy transvenous atrial defibrillation in patients with chronic atrial fibrillation. PACE 1997;20:899–90.
- Murgatroyd FD,
- Johnson EE,
- Cooper RA,
- et al.
- Levy S,
- Ricard P,
- Lau C-P,
- et al.
- Cooper RAS,
- Alferness CA,
- Smith WM,
- Ideker RE
- Tacker WA,
- Schoenlein WE,
- Janas W,
- Ayers GM,
- Wagner DO
- Keane D,
- Boyd E,
- Anderson D,
- et al.
- Levy S,
- Lauribe P,
- Dolla E,
- et al.
- Ayers GM,
- Ilina MI,
- Wagner DO,
- Sirokman WA,
- Griffin JC,
- Alferness CA
- Jung W,
- Tebbenjohanns J,
- Wollpert C,
- Luderitz B