Author + information
- Received January 5, 1998
- Revision received July 29, 1998
- Accepted August 20, 1998
- Published online December 1, 1998.
- Luis A. Pagan-Carlo, MDa,
- John J. Allan, MDa,
- Kirk T. Spencer, MDa,
- Clay L. Birkett, BSEEa,
- Rick Myers, BSEEa and
- Richard E. Kerber, MD, FACCa,* ()
- ↵*Address for correspondence: Dr. Kerber, Department of Internal Medicine, University of Iowa Hospital, 200 Hawkins Drive, Iowa City, Iowa 52242
Objectives. This study was performed to determine the efficacy of new encircling overlapping multipulse, multipathway waveforms for transthoracic defibrillation.
Background. Alternative waveforms for transthoracic defibrillation may improve shock success.
Methods. First, we determined the shock success achieved by three different waveforms at varying energies (18–150 J) in 21 mongrel dogs after short-duration ventricular fibrillation. The waveforms tested included the traditional damped sinusoidal waveform, a single pathway biphasic waveform, and a new encircling overlapping multipulse waveform delivered from six electrode pads oriented circumferentially. Second, in 11 swine we compared the efficacy of encircling overlapping multipulse shocks given from six electrode pads and three capacitors versus encircling overlapping shocks given from a device utilizing three electrodes and one capacitor.
Results. In the first experiment, the encircling overlapping waveform performed significantly better than biphasic and damped sinusoidal waveforms at lower energies. The shock success rate of the overlapping waveform (six pads) ranged from 67 ± 4% (at 18–49 J energy) to 99 ± 3% at ≥150 J; at comparable energies biphasic waveform shock success ranged from 26 ± 5% (p < 0.01 vs. encircling overlapping waveforms) to 99 ± 5% (p = NS). Damped sinusoidal waveform shock success ranged from 4 ± 1% (p < 0.01 vs. encircling overlapping waveform) to 73 ± 9% (p = NS). In the second experiment the three electrode pads, one capacitor encircling waveform achieved shock success rates comparable with the six-pad, three-capacitor waveform; at 18–49 J, success rates were 45 ± 15% versus 57 ± 12%, respectively (p = NS). At 100 J, success rates for both were 100%.
Conclusions. We conclude that encircling overlapping multipulse multipathway waveforms facilitate transthoracic defibrillation at low energies. These waveforms can be generated from a device that requires only three electrodes and one capacitor.
For the past three decades the damped sinusoidal waveform (DS) has been the most widely used waveform to deliver current in transthoracic defibrillation. However, with the advent of internal defibrillators, alternative waveforms for defibrillation were introduced (1–4). Specifically, biphasic, sequential and multipulse shocks have been reported to be superior for internal and transthoracic defibrillation in animals and humans (1–9).
Multipulse, multipathway shocks could be superior for the termination of ventricular fibrillation (VF) for a number of reasons. An adequate intracardiac current flow might be achieved in a greater portion of the ventricular myocardium (9–12). Depolarization of cardiac fibers and myocytes, which are directionally sensitive to electrical field stimulation (13–15), might be facilitated. Improved access might be gained to the cell populations in the ventricle that are in various electrical states of polarization, hyperpolarization and recovery, thereby depolarizing more of them and terminating the re-entrant pathways of ventricular fibrillation.
We hypothesized that a multipulse, multipathway waveform, configured to generate a series of electrical vectors that would rapidly and completely encircle the chest in an overlapping fashion, would optimize defibrillation by best accomplishing all the above considerations. The purpose of this study was to compare the efficacy for transthoracic defibrillation of encircling overlapping multipulse, multipathway waveforms versus biphasic and damped sinusoidal single pathway waveforms.
This investigation was approved by the University of Iowa Animal Care and Use Committee. Studies were performed on 21 adult mongrel dogs with body weights ranging from 20 to 30 kg (Study 1) and on 11 adult swine of 16–21 kg body weight (Study 2). We performed the second study on swine rather than dogs to show that an overlapping waveform’s efficacy was not limited to a single chest configuration, but was effective on a different species with a different chest configuration. Anesthesia was induced with IV Fentanyl/Droperidol and sodium pentobarbital; after endotracheal intubation, ventilation was maintained with a Harvard respirator. Anesthesia was maintained by periodic supplemental IV administration of sodium pentobarbital. Arterial pressure and heart rate were monitored throughout the experiment. Arterial blood gases were checked periodically, and ventilation was adjusted to maintain physiologic parameters.
To induce ventricular fibrillation the internal jugular vein was isolated and a bipolar pacing electrode was introduced into the right ventricle; 60 Hz alternating current (5–10 V) was delivered down the electrode for 5 s. Confirmation of VF was obtained by ECG and blood pressure recordings. All defibrillating shocks were delivered after 30 s of ventricular fibrillation, at end-exhalation.
Electrodes and pathways
Study 1 (dogs)
Six self-adhesive pad-paired electrodes (surface area 25 cm2each; Hewlett-Packard Corporation, McMinnville, Oregon), constructed of a conductive-adhesive polymer with a nonconductive backing were used. Firm electrode to chest wall pressure was assured by placing the electrodes under an elastic Velcro chest wrap. The electrodes were placed on a closely shaven chest wall encircling the chest in a single axial plane, equally spaced. Three current pathways resulted from this configuration (Fig. 1). Once placed, the pads were never moved.
Study 2 (swine)
Three pads were placed in a sagittal plane, equally spaced; two of the electrodes were on the lateral chest and the third was placed on the sternum. Three different current pathways were created by this configuration (Fig. 2).
The defibrillation shocks were delivered using three modified Hewlett-Packard defibrillators (Hewlett-Packard Corp.), each with a 242-μF capacitor. The defibrillators were modified to deliver truncated exponential shocks; the timing and polarity of the shocks were controlled by a Hewlett Packard Arbitrary Waveform generator (HP 33120 A) attached to each defibrillator. The waveform generators were programmed using a personal computer connected by an IEEE-488 bus. A computer program allowed selection of the number of pulses, polarity, duration and timing. The voltage of the shocks was set using a control on each defibrillator.
We planned the experiments so that each waveform tested would deliver the approximate equivalent of four energies: 25, 50, 100, 150 J. The voltage of each capacitor was adjusted to approximate the desired energy (e.g., 25 to 150 J), but we were limited to preset voltage increments and could not achieve the exact voltages desired. Thus, we have grouped the calculated energies into energy ranges (18–49, 50–99, 100–149, 150–200 J) rather than specific energies. The component pulse widths of each waveform were chosen before the study began and were not adjusted to achieve desired energies.
Damped sinusoidal waveform shocks were given from a standard Hewlett-Packard “Codemaster XE” defibrillator. For the damped sinusoidal waveform shocks, current and impedance data were inadvertently not recorded; we have provided data from previous studies in our laboratory using the same defibrillator and similar sized dogs (16).
A single modified defibrillator with a 242-μF capacitor was used to generate the encircling overlapping multipulse waveform (EO) shocks given from three electrode pads, while the three defibrillators (three capacitors) used in Study 1 were also used here to generate the EO shocks given from six electrode pads.
The formulas for the calculation of delivered energy from the voltages are given in the Appendix.
Three shock waveforms were evaluated. Biphasic waveform shocks and damped sinusoidal shocks were delivered along a single pathway, pathway 2 (Fig. 1). This pathway was chosen based on preliminary experiments that showed that the three pathways had similar impedances but pathway 2 tended to have the lowest impedance. Encircling overlapping shocks were delivered over three pathways (Fig. 1).
The EO and biphasic waveform in our experiments were all variations of a truncated exponential waveform, with about 25% total tilt. The biphasic waveform consisted of an initial 5-ms positive pulse followed by a 1-ms negative pulse, separated by 0.1-ms delay, delivered along the same pathway. In all multipulse multipathway shocks three capacitors were used and the leading edge voltage of each pulse was identical. In the biphasic single pathway shocks one capacitor was used, and the leading edge voltage of the negative component was equal to the trailing edge voltage of the positive component.
The EO waveform in this study was constructed by discharging one capacitor followed by a second capacitor discharge before the end of the first pulse. The sequential discharges resulted in a current vector that encircled the chest in a counterclockwise manner resulting in 11 electrical vectors. At any one time, one or two capacitors were discharging either simultaneously or in sequence. Figure 1shows the electrode placement and resultant electrical vectors: vector a was the result of capacitor 1 discharging, which generated a pulse traveling between electrodes 1 and 4; vector b was the result of capacitors 1 and 2 discharging simultaneously (capacitor 2 discharge begins 0.64 ms after capacitor 1 discharge begins), which generated simultaneous pulses between electrodes 1–4 and 2–5, respectively; vector c was the result of capacitor 2 discharging (capacitor 1 off), which generated a pulse between 2 and 5 only; vector d was the result from capacitor 2 and 3 discharging simultaneously, which generated simultaneous pulses between electrodes 2–5 and 3–6, etc. This sequence was repeated until the encircling current pulses were completed. The result was a current vector that traveled from pad 1 to 6 in a counterclockwise manner, creating all the 11 vectors (e–k) shown in Figure 1. Pulses 1 and 6 are shorter than pulses 2–5; the duration of each electrical vector is equal, approximately 0.64 ms. The total shock duration was approximately 7 ms.
The EO waveforms in Study 2 were constructed as shown in Figure 2. Using three electrodes and a single capacitor, or six electrodes and three capacitors, seven vector (a–g) encircling overlapping multipulse shocks were created. In the three-electrode configuration current flowed from one anode (+) to either one cathode (−) or simultaneously to two cathodes as the multipulse shock continued. The total duration of the shock was 7 ms, from both the three-electrode and six-electrode system. The three-capacitor, six-electrode EO waveform in this study was similar to that used in Study 1, except that some of the intermediate vectors were deleted, so that each pulse was 2 ms in duration, and a total of seven rather than 11 directional vectors were created during the multipulse shock. For comparison, damped sinusoidal and biphasic waveforms were also tested. The biphasic waveform used in this study was comprised of an initial 5-ms positive pulse followed by a 3-ms negative pulse. The damped sinusoidal waveform was identical to that used in Study 1.
In addition to the damped sinusoidal, biphasic and short-duration EOs discussed above, we also evaluated other waveforms but have not included the data in this paper: monophasic (trapezoidal) waveforms, encircling nonoverlapping waveforms and long-duration encircling overlapping waveforms. A total of 17 different waveform configurations of varying durations, overlap and symmetry were evaluated in Study 1. In Study 2, in addition to the four waveforms reported, single-capacitor three-electrode-encircling overlapping waveforms of 6- and 10-ms duration were also studied. Data from these additional waveforms are not included in this report. To keep the experiments a reasonable length, each animal typically received shocks from only three to four waveforms, and the studies lasted 4–5 h. Hemodynamic deterioration was uncommon; if it did occur the study was terminated.
Preparation of the contact skin between electrodes was accomplished by delivering three preliminary shocks using a standard damped sinusoidal shock defibrillator, at an energy of 50 J, along each pathway before any data collection. This was done to minimize the effects of repeated shocks on transthoracic impedance (TTI), since the greatest TTI decline occurs after the initial shocks (17,18). Following 30 s of VF, one of the experimental shocks was delivered. If it failed to terminate VF, a damped sinusoidal shock of at least 200 J was delivered to “rescue” the dog; data from such “rescue” shocks were not used to calculate waveform efficacy. The animal was allowed to return to preshock arterial pressure before the next fibrillation/defibrillation episode; the minimum time between VF episodes was 3 min. This sequence was repeated four times at each energy. The energies fell within the following ranges, 18–49, 50–99, 100–149, 150–200 J. The results of the four shocks at each energy were averaged to yield one data point. The percent success of each shock in terminating VF was calculated. In both experiments 1 and 2, the sequence of waveforms studied in each animal was random, and the energy sequence within each waveform was random. Once a waveform and energy were chosen, all four shocks were delivered at that energy.
Proportion success data obtained from 21 dogs in a factorial arrangement of three waveforms and four energy level classes were fit by a mixed model analysis of variance (ANOVA), with the dog as the random variance source, and waveform, energy level class and their interaction as the treatment sources of variation. The model accounted for 53.7% of the variation in proportion success. Multiple comparisons involved Tukey’s adjustment (19). All results are presented as mean ± SE.
Friedman’s test was used to compare percent shock success among the four waveforms at each energy level. Post-hoc testing involving Friedman mean rank scores was performed for pairwise comparison between waveforms. The p values from the post-hoc test have been adjusted by the total number of comparisons. Adjusted p < 0.05 was considered statistically significant. All results are presented as mean ± SE.
In Table 1the percent success rate of the three waveforms are presented. The encircling overlapping multipulse waveform shocks achieved significantly (p < 0.01) higher success rates than both biphasic and damped sinusoidal waveform shocks at energy ranges of 18–49 and 50–99 J; EO success rates were 67 ± 4% at 18–49 J and 89 ± 4% at 50–99 J. At 100–149 J the encircling overlapping shocks were superior to damped sinusoidal shocks at 18–49 and 100–149 J. These results are highlighted in Figure 3.
Table 2presents the percent shock success of the four waveforms tested: encircling overlapping shocks from six electrode pads (three capacitors), encircling overlapping shocks from three electrode pads (one capacitor), biphasic shocks and damped sinusoidal shocks. The EO six-electrode, three-capacitor shock success ranged from 57 ± 12% at 18–49 J to 100% at >150 J. The EO three-electrode, one-capacitor success rates ranged from 45 ± 15% to 100%, and were not significantly different from the six-electrode, three-capacitor success rates at any energy. Biphasic shock success ranged from 7 ± 7% to 100%, and was less than the EO six-electrode, three-capacitor waveform at the lowest energy range. Damped sinusoidal shock success ranged from 0% to 89 ± 7%, and was lower than both the encircling overlapping waveforms at all energy ranges except >150 J, and lower than the biphasic waveform success at 50–99 J and 100–145 J. These results are highlighted in Figure 4.
The main conclusions of these studies are that: 1) at low energies, encircling overlapping multipulse, multipathway waveforms are superior to single-pathway biphasic and damped sinusoidal waveforms for transthoracic defibrillation. At higher energies, this superiority was not maintained; and 2) effective encircling overlapping waveforms can be generated from a three-electrode single-capacitor defibrillator, a potentially clinically applicable device.
Why should an overlapping encircling multipulse shock waveform facilitate defibrillation?
Myocardial fiber orientation in relation to a defibrillating or stimulating intracardiac current has been shown to be important in achieving depolarization of the cell. Individual myocytes are easier to stimulate and depolarize when the cell is parallel to the stimulating field than when it is orthogonal to it (13–15). Roberts et al. (20)have shown that cardiac fiber orientation influences myocardial conduction velocity and tissue resistivity. Thus, the orientation of the delivered transthoracic current may play a role in achieving defibrillation. A single-pathway electrical pulse may depolarize those myocytes oriented parallel to the electrical field (the optimum orientation) but not perpendicular to the field. In contrast, the rapidly shifting electrical vector achieved by an overlapping multipulse waveform (Figs. 1 and 2)may facilitate depolarization of a larger population of myocytes of varying orientations. These explanations implicitly assume that the electrical field is uniform throughout the ventricles for each pathway, and that the depolarizing effect of a particular electrical field is the same for all cells that are oriented at a particular angle to that field.
We have also previously shown that overlapping multipulse, multipathway shocks achieve higher intracardiac current flow during the overlap phase (8).
Applying these concepts, we have previously shown that sequential overlapping dual-pulse shock waveforms delivered over two orthogonal pathways facilitate transthoracic defibrillation compared with a monophasic truncated exponential waveform, and that the direction of the net electrical vector changes during such overlapping pulses (8). The present investigation extends this concept to a completely encircling multipulse waveform.
Potential clinical applications
In our second study, we showed that encircling overlapping waveforms can be generated from a three-electrode, single-capacitor defibrillator; shocks from this device were as effective as those from a more complex six-electrode, three-capacitor device, and at low energies (but not higher energies) maintained their superiority over damped sinusoidal and biphasic waveform shocks. This is important in considering potential clinical applications of this approach; a multipulse, multipathway waveform would have little clinical use if it required a large, heavy defibrillator and many electrodes. The three electrodes used in experiment 2 could be preincorporated into an elasticized belt with Velcro closures similar to what we used in this study; such a belt could be rapidly wrapped around a collapsed patient (or the patient could be rolled onto the belt). Since only a single capacitor is needed, the defibrillator size and weight could be reduced.
It has been shown in animals (21–25)and patients (2–4)that single-pathway biphasic waveforms are superior to the standard damped sinusoidal monophasic waveforms for transthoracic defibrillation. The apparent ability of this waveform to achieve equivalent defibrillation efficacy at lower energies has permitted the introduction of smaller and lighter defibrillators for prehospital use; such units are especially suitable for “public access defibrillation” (26). The overlapping encircling multipulse approach we have evaluated in these studies demonstrated superiority over single-pathway biphasic waveforms at low energies. However, the three electrodes required will necessarily increase the electrical complexity of the defibrillator to some degree. Whether a multipulse multipathway defibrillator for clinical use could be constructed of a size, weight and cost similar to the new single-pathway biphasic defibrillators remains to be determined.
There have been several anecdotal reports of successful human transthoracic defibrillation with multipulse/multipathway shocks (5,6). In those reports the shocks were delivered through two separate lead systems, and were temporally separated by several seconds. Although encouraging, those results cannot be directly extrapolated to our approach since our encircling shocks were continuous.
There are several limitations of our studies. First, only 30 s of VF was allowed before a shock was delivered. In the usual clinical scenario, VF occurs for much longer periods of time before the first shock; the efficacy of the waveforms demonstrated in this experiment may not be as high in the clinical setting after prolonged periods of fibrillation. Second, our defibrillator did not allow us to adjust the tilt of the waveforms generated to maximize their effects. Several in vivo studies, as well as mathematical models, have shown that tilt can play a significant role in maximizing defibrillation (27–29). Third, even though we showed that the overlapping encircling waveform is effective in two different species with different chest configurations, there are still differences compared with humans; the canine and swine hearts are more medially located within the thorax than in the human heart. Whether this difference would alter the effectiveness of encircling waveforms for defibrillation remains to be determined.
In summary, encircling overlapping multipulse multipathway shock waveforms were superior to single-pathway waveforms at low energies, but not at higher energies. Further evaluation of these waveforms are appropriate.
We gratefully acknowledge the review and criticisms of Drs. Janice Jones and Michael Kallok, the statistical assistance of Carl K. Brown and Miriam B. Zimmerman, the technical assistance of Robin A. Smith, BA and the secretarial assistance of Diane Phillips.
The defibrillation shocks were measured using an ISC-16 digitizing board and EGAA software provided by R.C. Electronics. The data acquisition board was configured for ±10 V, single ended. The sample rate was 50 kHz per channel.
The following terms were used in the calculation of delivered energy: Edel= Energy delivered to load; Esto= Energy stored on capacitor; Vi= Initial voltage on capacitor at start of discharge; Vf= Final voltage on capacitor at end of discharge; VL= Leading edge voltage, voltage across the chest; IL= Leading edge current, current through the chest; RL= Pathway impedance (chest impedance), calculated from VL/IL; RI= Internal impedance (Capacitor ESR, wiring resistance, etc.); C = Energy storage capacitor
Energy stored on a capacitor is Esto= 1/2CV2. Energy removed from the capacitor during a discharge can then be found from 1/2C(VI2− Vf2). Then, energy delivered to the load is: Edel= 1/2C(VI2− Vf2)·RL(RL+ RI).
Study 1.In study 1, the voltage on the capacitor (VIand Vf) was not measured, but can be calculated from the following: VI= VL·(RL+ RI)/RL, where RL= VI/IL. Vfcan then be calculated from Vf− Vie−t(RC), where tis the width of the discharge pulse (seconds) and R = RI+ RI. Combining the above equations leads to the following formula for calculating delivered energy: The term t(time) in the above equation is the time a given pathway is on. For example, using a biphasic waveform with a positive phase of 5 ms, and a negative phase of 3 ms, twould be 8 ms.
The energy obtained using the above formula was multiplied by 3 if an encircling waveform was used. The manufacturer of the defibrillator (Hewlett-Packard Corp.) tested the pulse characteristics of the three defibrillators before the beginning of the study to determine the pulse characteristics of the three were the same.
The term 1.07 is a correction factor to account for the underestimation of the actual voltage measurement. Recalibration of the equipment at the end of the study demonstrated our digitizing board underestimated the voltage and current measurement by approximately 7%. The correction term 1.07 in the formula corrects for this error.
Study 2.A single modified defibrillator with a 242 μF capacitor was used to generate the EO shocks from three electrode pads, while the three defibrillators (three capacitors) used in Study 1 were also used here to generate the EO shocks given from six electrode pads. The test system was modified to allow a direct measurement of the voltage on the energy storage capacitor. Then, delivered energy was calculated using the following formula:
☆ This work was supported in part by a grant from the Hewlett-Packard Corp. (Andover, Massachusetts), by a grant from the Laerdal Foundation for Acute Medicine (Stavanger, Norway), by an Institutional National Research Service Award (HL07121) (LP-C), and an award from the National Heart Lung and Blood Institute (NHLBI) (HL53284) (REK).
- damped sinusoidal waveform
- encircling overlapping multipulse waveform
- ventricular fibrillation
- Received January 5, 1998.
- Revision received July 29, 1998.
- Accepted August 20, 1998.
- American College of Cardiology
- Jones J.L.
- Bardy G.H.,
- Gliner B.E.,
- Kudenchuk P.J.,
- et al.
- Bardy G.H.,
- Marchlinsky F.E.,
- Sharma A.D.,
- et al.
- Hoch D.H.,
- Batsford W.P.,
- Greenberg S.M.,
- et al.
- Chang M.S.,
- Inoue H.,
- Kallok M.J.,
- Zipes D.P.
- Kerber R.E.,
- Spencer K.T.,
- Kallok M.J.,
- et al.
- Jones D.L.,
- Klein J.G.,
- Guiraudon G.M.,
- et al.
- Chen P.S.,
- Shibata N.,
- Dixon E.G.,
- et al.
- Frazier D.W.,
- Krassowska W.,
- Chen P.S.,
- et al.
- Tung I.,
- Sliz N.,
- Mulligan M.R.
- Chen P.S.,
- Cha Y.M.,
- Peters B.B.,
- Chen L.S.
- Geddes L.A.,
- Tacker W.A.,
- Cabler P.,
- Chapman R.,
- Rivera R.,
- Kidder H.
- Dahl C.F.,
- Ewy G.A.,
- Ewy M.D.,
- Thomas E.D.
- ↵SAS Institute Inc., SAS/STAT software. Changes and Enhancements Through Release Grill, Cary NC: SAS Institute Inc., 1996;110H pp GLM:317–24.
- Roberts D.E.,
- Hersh L.T.,
- Scher A.M.
- Gliner B.E.,
- Lyster T.E.,
- Dillion S.M.,
- Bardy G.H.
- Dixon E.G.,
- Tang A.S.L.,
- Wolf P.D.,
- et al.
- Jones J.L.,
- Swartz J.F.,
- Jones R.E.,
- Fletcher R.
- Weisfeld M.,
- Kerber R.E.,
- McGoldrick P.R.,
- et al.
- Irnich W.