Background Measurements of refractory period extension by shocks during ventricular pacing at fast rates predict that all tissue should be refractory for a brief interval after shocks during fibrillation. This study experimentally determined whether a refractory interval was present just after a shock during fibrillation.
Methods and Results In pentobarbital-anesthetized dogs, rectangular monophasic (4-ms) or biphasic (2.5/1.5-ms) shocks were followed with a 2-ms postshock stimulus (PSS) delivered to the defibrillation electrodes. We measured the effect of PSS on the shock current (I50) required for 50% defibrillation success. In group 1 (n=6), a 1.0-A PSS had no effect on I50 when delivered up to 35 ms after monophasic shocks but greatly increased I50 when delivered at 50 to 90 ms. A 0.5-A PSS had no effect at any timing. In group 2 (n=6), we compared 1.0-A PSSs after monophasic and biphasic shocks. The effect of PSS after monophasic shocks was similar to group 1. After biphasic shocks, PSS at the same timings had similar effects but caused even greater increases in I50.
Conclusions We conclude that after both monophasic and biphasic shocks during fibrillation, there is a postshock interval during which the heart is refractory to the refibrillating effect of PSS. The interval is shorter for biphasic than for monophasic shocks with the same duration and defibrillation efficacy. These findings support the refractory period extension hypothesis for defibrillation and suggest that propagating depolarization activity is absent immediately after defibrillation shocks but that it develops again at the end of the refractory interval or later.
It is generally accepted that electrical defibrillation involves the direct excitation of myocardial tissue that is capable of being excited at the instant of the shock. However, in vivo1 2 3 4 5 6 7 8 and in vitro9 10 11 studies of the effect of a shock on refractory period and cellular action potential suggest that successful defibrillation also involves a shock-induced extension of refractory period (or action potential). Such studies have shown the shock effect to depend on both the intensity and timing during the refractory period. A greater RPE occurs for stronger shocks or shocks that occur later during the refractory period.
VF is thought to be composed of reentrant activation wave fronts that move throughout the myocardium in a disordered pattern. If this view of VF is valid, then at any instant, different locations in the myocardium must be at different refractory timings in accordance with the pattern of the reentrant pathways. Since a shock occurs in all tissue simultaneously, it must therefore encounter tissue with various refractory timings. Thus, the same shock could (1) directly excite tissue at locations along the reentrant pathways at which tissue was recovered, (2) cause RPE at locations along the pathways at which tissue was later in its refractory period, and (3) have little effect at locations along the pathways at which tissue was early in its refractory period.
In keeping with this idea, shocks during fibrillation have been predicted to produce an interval after the shock during which all tissue is refractory to further excitation, according to the following reasoning.1 2 Tissue that was directly excited by the shock would remain refractory while it undergoes a new, albeit premature, refractory period. Tissue that was late in its refractory period at the instant of the shock would experience RPE and thus be delayed from recovering quickly after the shock. Tissue that was early in its refractory period would experience little RPE but would remain refractory until the end of its unextended refractory period. This predicted refractory interval after the shock has been hypothesized to contribute to defibrillation by preventing propagation immediately after the shock.
Although these actions of defibrillation shocks on fibrillating myocardial tissue are conceptually straightforward, their direct experimental observation has remained elusive. RPE has been relatively well characterized in in vivo preparations1 2 3 5 6 7 8 during ventricular paced rhythms under conditions of rate8 and ischemia2 that mimic VF. However, RPE cannot be measured directly during fibrillation. On the other hand, action potentials from fibrillating myocardial epicardium of several species have been observed by use of monophasic action potential,5 optical,9 and direct cellular recording4 methods. Although those observations show that shocks during VF extend cellular action potentials, their technical complexities have prevented action potential prolongation during VF from being well characterized.
Since tissue is rarely fully recovered12 during fibrillation, much (if not most) of the fibrillating tissue is early in its refractory period at the instant of the shock; therefore, its recovery timing should not be greatly affected by the shock. Accordingly, it has been predicted that the pattern of the fibrillatory reentrant pathways should be largely “remembered” across the shock event.1 As a result, even after a successful shock, the fibrillatory pattern would be embedded in this tissue as a pattern of cellular recovery timing. This tissue should recover in approximately the same pattern with which it was excited by the previous fibrillatory activation wave front. Thus, any propagating postshock activity should be shepherded into approximately fibrillatory pathways.
We reasoned that even after a highly successful defibrillation shock, a PSS could directly excite tissue as it recovered along the partially “remembered” pathways. The result would be to reintroduce propagating depolarization activity into the remembered pattern of fibrillation. Thus, a highly successful shock could be turned into a highly unsuccessful shock by the presence of a postshock pacing stimulus. On the other hand, if all tissue were refractory to stimulation by the PSS, then it could not reexcite the reentrant pathways and thus would have little effect on defibrillation efficacy.
The main objective of this study was to experimentally test whether a refractory interval existed just after a defibrillation shock by testing the ability of a PSS to lower defibrillation efficacy. A second objective of the study was to compare the effect of a PSS after monophasic versus biphasic shocks to determine whether improved biphasic defibrillation efficacy was related to a change in the postshock environment.
Healthy adult mongrel dogs were anesthetized with sodium pentobarbital (35 mg/kg induction, ≈4 mg·kg−1·h−1 maintenance) and ventilated with room air. The right femoral artery and vein were cannulated to measure arterial pressure and to deliver fluids. Arterial blood gases were monitored (IL 1304, Instrumentation Laboratories) and ventilation was adjusted throughout the experiment to maintain normal values. Via a left thoracotomy, stainless steel disk defibrillation electrodes (25 mm in diameter) were positioned on the left ventricular apex (cathode) and right ventricular base. A bipolar pacing electrode was positioned in the right ventricular outflow tract for induction into fibrillation. After electrode placement, the pericardium was closed, the ribs were approximated, and the skin was closed. The ECG (lead II) and arterial pressure were monitored to confirm fibrillation induction and termination. Our institution is accredited by the American Association for Accreditation of Laboratory Animal Care, and these experiments conformed to their guidelines for the use of animals in research and were approved by the institution's animal use committee.
Experimental protocols involved multiple measurements of the defibrillation threshold as the current (and energy) required for 50% success at defibrillation with various waveforms. Defibrillation shocks were delivered by a custom arbitrary waveform defibrillator capable of producing arbitrary current-controlled waveforms. The voltage of each shock was observed on a storage oscilloscope to confirm the waveform shape and to determine resistance.
Animals were induced into VF with a 50-Hz train of 2-ms stimuli (20 to 25 mA) to the pacing electrode. Defibrillation was always attempted 15 seconds after the onset of VF. A rescue shock was used immediately (if needed), and fibrillation/defibrillation episodes were repeated at 3-minute intervals. The I50 was determined with a previously reported13 three-reversal up/down protocol. A recovery period of ≥6 minutes was provided between defibrillation threshold determinations.
In this preparation, typical I50s for the 4-ms rectangular shocks are about 4 A, but PSSs could increase this value. Because of concerns with shock damage to the tissue and the output capability of the defibrillator, we decided to limit shocks to 8 A (approximately twice the normal I50). If an 8-A shock did not defibrillate, 8 A was used as the I50.
Group 1: Effect of PSS After Monophasic Shocks
In group 1 (27.7±4.1 kg, n=6), we examined the effect of a small PSS delivered directly to the shock electrodes. The defibrillation shock had a 4-ms monophasic rectangular waveform. I50s were determined for these shocks when they were followed by a 2-ms rectangular PSS given 5, 20, 35, 50, 70, or 90 ms after the trailing edge of the shock (random order). The intensity of the PSS was the same for the entire set of measurements and was either 0.5 or 1.0 A (random). The set of measurements was then repeated using the other postshock intensity and a new random order for PSS timings. At the start, middle, and end of the experiment, control I50s were measured with only the 4-ms rectangular shock. The average of these control values was used to normalize the other data from each animal.
Group 2: Effect of PSS After Monophasic Versus Biphasic Shocks
In group 2 (26.1±4.0 kg, n=6), we compared the influence of the PSS on monophasic and biphasic waveforms. The monophasic shock had a 4-ms rectangular waveform. The biphasic shock had a 2.5/1.5-ms rectangular waveform, with both phases having the same amplitude. The I50s for each of these waveforms were measured as controls (random order) at the start and end of the protocol. Similar to group 1, I50s were measured when these waveforms were followed with a 1.0-A PSS (2 ms rectangular) delivered 10, 25, 40, 55, and 70 ms after the shock. The order of these timings was random. At each timing, both the monophasic and biphasic I50s were determined in random order. As in group 1, the average of the appropriate control I50s was used to normalize the other data from each animal.
Data Analysis and Statistics
Data are reported as mean±SD. Statistical tests were considered significant at the .05 level and were performed with commercially available statistical analysis software (JMP version 3.02, SAS Institute). Statistical tests are described as their outcomes are reported.
Group 1: Effect of PSS After Monophasic Shocks
The 4-ms rectangular shocks required 4.31±0.77 A and 5.96±2.12 J to defibrillate with 50% efficacy. In comparison, the 0.5- and 1.0-A pulse-shock stimuli were very small and delivered only about 0.04 and 0.16 J, respectively. The effects of these PSSs are shown in Fig 1⇓, which plots the normalized current for 50% success against the PSS timing. The 1.0-A PSS had a profound effect on defibrillation efficacy when delivered 55, 70, or 90 ms after the shock. Yet, the same 1.0-A stimulus had no significant effect when delivered 5, 20, or 35 ms after the shock. The elevated I50s shown in Fig 1⇓ for the 1.0-A stimulus are conservative, because the I50 was too high to measure in most animals when the stimulus was given at 70 ms and in several animals when the stimulus was delivered at 50 or 90 ms after the shock. The 0.5-A PSS has no statistically significant effect on defibrillation efficacy at any timing. However, the 0.5-A PSS delivered at 70 ms just failed (P=.063) to significantly increase the I50.
The electrical resistance fell from 92 to 69 Ω over the course of the experiment (ANOVA for linear fit), which is typical for this experimental preparation. However, there was no significant difference among the resistances (one-way ANOVA) for the different I50 determinations.
Group 2: Effect of PSS After Monophasic Versus Biphasic Shocks
In group 2, the 4-ms rectangular shocks required 4.12±0.73 A and 5.54±1.82 J to defibrillate with 50% efficacy. The 4-ms biphasic controls required only 3.65±0.40 A and 4.33±0.96 J to defibrillate, demonstrating that there was a strong biphasic effect (paired t test). The effects of the PSSs are shown in Fig 2⇓, which plots the normalized I50 against the PSS timing. The data in Fig 2⇓ are normalized by the average of the control shocks of the same type. Thus, for both shock types, a value of 1 indicates that the PSS had no effect on I50.
As in group 1, the 1.0-A PSS had no significant effect on I50 when delivered 10 or 25 ms after monophasic defibrillation shocks. However, the same stimulus significantly increased I50 when delivered 40, 55, and 70 ms after the shock. When delivered 10 or 25 ms after the shock, the PSSs had no significant difference after biphasic versus monophasic waveforms (paired t test). However, when delivered 40, 55, or 70 ms after the shock, the PSS increased I50 significantly more after biphasic than after monophasic shocks. Thus, after biphasic shocks, PSSs increased I50 after shorter postshock delays.
As in group 1, the electrical resistance for group 2 fell from 94 to 70 Ω over the course of the experiment (ANOVA for linear fit), and there was no significant difference among the resistances (one-way ANOVA) for the different I50 determinations.
Refractory Interval After the Shock
This study demonstrated that a PSS could profoundly reduce defibrillation efficacy when delivered >35 ms after monophasic shocks or >25 ms after biphasic shocks. But the same PSS had virtually no effect on defibrillation efficacy when given earlier. One plausible explanation for this behavior is that all myocardial tissue was refractory to stimulation for up to 35 ms after monophasic shocks (25 ms after biphasic), so that the 1.0-A PSS failed to have an effect. When delivered later, tissue would be more recovered after the shocks, so that the PSS could excite tissue and restart the fibrillation event.
Direct in vivo measurements of RPE during ventricular pacing at rates up to those associated with fibrillation2 8 have demonstrated the existence of a 35- to 45-ms (depending on shock intensity) postshock refractory interval. During that interval, a pacing stimulus could not evoke a propagated response. The present study demonstrates a similar interval after shocks during fibrillation during which tissue is refractory to the refibrillatory effect of a postshock field stimulus.
Referring to the postshock interval as a “refractory interval” must be undertaken with some caution. The term “refractory” usually implies that stimulus failed to produce a propagated activation. However, from the data in this study, we cannot be certain whether early PSSs failed to evoke propagated responses or whether they evoked responses that subsequently failed to affect defibrillation efficacy. Thus, in this study, “refractory” means that the PSS failed to affect defibrillation efficacy.
As proposed in the introduction, it could be that the PSSs restarted the fibrillation episode by directly exciting tissue that recovered first after the otherwise successful shock. In this scenario, the stimulus would evoke activations that were then forced to propagate along at least some portions of the fibrillatory pathways that were “remembered” from before the shock. Accordingly, the evoked responses would have a high likelihood of refibrillation, thus lowering efficacy by allowing otherwise successful shocks to be unsuccessful. Previous reports that the first activations after unsuccessful shocks follow a pattern that is similar to the one just before the shock14 are consistent with this possibility.
In keeping with this scenario, a PSS given after all tissue had recovered (about 1 cycle of fibrillation) would be predicted to have little or no refibrillating effect, because few or none of the remembered pathways would be present. Although we did not introduce PSSs >90 ms after the shock, Fig 1⇑ shows that the degrading effect of the PSS appeared to be reduced with increased timing past 70 ms.
The failure of the 0.5-A stimuli to affect efficacy could be the result of a PSS strength-interval relationship. To evoke a response, the weaker PSS might require additional postshock delay for tissue to become more fully recovered. However, with the additional delay, less of each remembered pathway would be present, so that the evoked responses would be less likely to renew the fibrillation.
Instead of restarting the fibrillation event, it is also possible that the PSSs induced new fibrillation events in the same way that they could if delivered during a ventricular paced rhythm. Both the 1.0- and 0.5-A stimuli are sufficient to induce fibrillation when delivered during the vulnerable period during paced ventricular rhythm. From the data in this study, it is not possible to distinguish between these possibilities.
Comparison With Other Works: Isoelectric Window
Previous experimental studies15 16 17 18 have found that no activations are present during an “isoelectric window” after either successful or unsuccessful shocks during fibrillation. By delivering a defibrillation shock to the heart and including an opposite-polarity shock to a small region of the myocardium, Zhou et al18 were able to control the site at which activations first appeared after defibrillation shocks. After monophasic shocks, they measured the isoelectric window to be from 56 ms (weakest intensity) to 89 ms (strongest intensity), depending on shock intensity at the early site. After biphasic shocks with the same intensities, the isoelectric windows were shorter (43 and 54 ms, respectively). In other studies from the same laboratory,17 isoelectric window duration after monophasic defibrillation was measured to be 42 ms for shocks near (100 V beneath) the defibrillation threshold. The 35-ms (monophasic) and 25-ms (biphasic) refractory intervals found in this study are in keeping with these isoelectric window durations.
Comparison With Other Works: Effect of Shock Intensity
Both the duration of the isoelectric window15 16 18 after defibrillation shocks and the minimum refractory interval2 8 19 after shocks during paced rhythm have been shown to vary with the intensity of the shock. The present study indirectly suggests a similar behavior after shocks during fibrillation. A PSS given 35 ms after the shock was within the refractory interval. When delivered only 15 ms later (50 ms after the shock), the PSS profoundly reduced defibrillation efficacy. However, it was still possible (in most cases) to increase the shock intensity to achieve a 50% level of success even with the PSS present. This suggests that increasing the shock strength produced a longer refractory interval so that 50-ms PSS was now within it or nearly within it.
Comparison With Other Works: Effect of Waveform
Previous studies have also demonstrated that compared with monophasic shocks, biphasic shocks produce shorter isoelectric windows,18 less action potential prolongation,3 4 and less RPE.20 21 The present study is consistent with these earlier studies because the biphasic curve in Fig 2⇑ appears to be shifted to the left from the monophasic curve. This suggests that the refractory interval after the biphasic shock is shorter. Of course, since biphasic waveforms required a lower I50, the leftward shift might also be due to a lower shock intensity. It is also possible that after a biphasic versus monophasic shock, the same PSS has a greater stimulatory effect and is thus better able to restart the fibrillation episode. Further studies that vary PSS timing and intensity would be required to distinguish between these possibilities.
Implications for Defibrillation
A defibrillation shock is normally thought of as an instantaneous event that either does or does not defibrillate the heart. However, this study suggests that defibrillation is not instantaneous but rather a process that takes place over some finite time. Decades of defibrillation research demonstrate that the size and shape of the shock govern defibrillation success, so the defibrillation process must start at the instant of the shock. However, as in group 2 of the present study, a PSS delivered 40 ms after the shock can have a different effect after monophasic than after biphasic shocks with the same efficacy. This shows that even 40 ms after the shock, there is some difference in the electrophysiological state of the myocardial tissue, suggesting that either the monophasic or the biphasic defibrillation is still in progress.
The demonstration of an interval after the shock during which a PSS does not alter defibrillation efficacy may also have some basic implications regarding the nature of defibrillation. During I50 measurements, shocks with both successful and unsuccessful outcomes were encountered with approximately equal frequency. This was true for control I50s and also for I50s measured while PSSs were given during the postshock refractory interval. If these early PSSs preferentially affected either successful or unsuccessful shock attempts, then the overall defibrillation efficacy would have been altered. However, it was not. There are several logical possibilities that could account for the failure of these PSSs to affect efficacy.
One possibility is that the postshock refractory interval followed only successful shocks and that activations continuously propagated after unsuccessful shocks. Then, a PSS would have no effect after either successful shocks (because tissue was refractory) or unsuccessful shocks (because activations were already present). Given the numerous studies that have demonstrated RPE,2 6 7 8 19 20 action potential prolongation,3 4 6 9 10 11 and postshock isoelectric windows3 15 16 17 18 to vary relatively smoothly with shock intensity, it seems unlikely that shocks of the same intensity delivered to the same tissue with the same electrodes could cause 35 to 40 ms of postshock refractoriness in some cases (successes) but little or none in others (failures). This possibility is also inconsistent with the observations of an isoelectric window after both successful and unsuccessful shocks. However, this possibility cannot be ruled out from data of the present study.
Another possibility is that activations propagated during the postshock interval after both successful and unsuccessful shocks. Then, a PSS during this interval would have little effect because activations were present in both cases. However, this possibility requires that the activations either terminated (successful) no sooner than 35 to 40 ms after the shock or continued (unsuccessful) into fibrillation according to some unknown factor. This possibility is also inconsistent with a postshock isoelectric window but cannot be ruled out from the data in the present study.
A third possibility is that PSSs during the postshock refractory interval had a carefully balanced effect to turn unsuccessful shocks into successful ones and successful shocks into unsuccessful ones. Although the net result would also be an unchanged defibrillation efficacy, this possibility also seems very unlikely.
A possibility that we consider the most likely is that no propagated activity was present during the postshock interval after both successful and unsuccessful shocks because no tissue was sufficiently recovered to support propagation. The PSS would then have no effect unless it was delivered after the postshock interval, when tissue had started to recover. This possibility is in agreement with the predictions made on the basis of measurements of RPE during ventricular pacing1 2 6 7 8 19 and observations of isoelectric windows and postshock activations after shocks during fibrillation.14 15 16 17 18
However, this last possibility poses a dilemma. If the 1-A electrical stimulus during the postshock refractory interval could not produce a propagated response, then it is reasonable to think that other forms of stimulation (including cell-to-cell excitation) should also fail to propagate. But if activations failed to propagate during the 35- to 40-ms postshock refractory interval, where did postshock activations arise after the unsuccessful shocks that comprised approximately half of the attempted defibrillation episodes?
This is essentially the same dilemma as encountered in earlier works15 16 17 18 that observed isoelectric windows after both successful and unsuccessful shocks. However, whereas the earlier studies suggested that activations were absent during the postshock interval, the present study enhances this dilemma by further suggesting that tissue is incapable of being excited during the postshock interval.
If this last possibility is valid, then the only resolution for the dilemma is that new activations are produced in the tissue after approximately half of the shocks at I50. Moreover, the activations could not arise until the end of the postshock refractory interval, when at least some tissue was capable of supporting propagation.
It is beyond the scope of the present study to explore possible mechanisms by which postshock activations could arise many milliseconds after the shock. However, on the basis of our previous studies of RPE, we have speculated on a possible mechanism19 that involves postshock electrotonic stimulation by sharpening spatial gradients of transmembrane potential.
At different locations along a reentrant pathway, the same shock could affect tissue differently, depending on the preshock state of refractoriness of the tissue. It has been found2 6 7 8 19 that tissue near (80%) the end of its refractory period when the shock was delivered recovers earliest after the shock, and tissue with other timings recovers later. This could produce spatial gradients of transmembrane potential in the tissue and thus sources for electrotonic stimulation. A simplified model based on RPE measurements in dogs predicts the spatial gradients to reach a maximum many (20 to 40) milliseconds after the shock. Also, the maximum gradient is predicted to be inversely related to shock intensity. The result would be activations that do not appear until well after the shock and that fail to appear at higher shock intensities.
An increased homogeneity of cellular recovery9 10 after the shock is normally considered to favor defibrillation by making postshock reentry less likely. The proposed mechanism, on the other hand, suggests that the increased homogeneity has a more direct effect in preventing postshock activations from arising by reducing the size of postshock spatial gradients. Thus, the concept of postshock electrotonic stimulation could account for many features of defibrillation. Although this concept remains speculative, experimental studies to assess its validity could be a productive area for future research.
We conclude that after both monophasic and biphasic defibrillation shocks, there is an interval of time during which the heart is refractory to the refibrillating effect of a PSS. The interval is shorter after biphasic shocks than monophasic shocks with the same duration and defibrillation efficacy. These findings support the refractory period extension hypothesis for defibrillation and suggest that defibrillation is not instantaneous but rather a process that may extend over many milliseconds.
Selected Abbreviations and Acronyms
|I50||=||current required for 50% success at defibrillation|
|RPE||=||refractory period extension|
This work was fully funded by the Lilly Research Laboratories Division of Eli Lilly and Company, Indianapolis, Ind. The authors are grateful for the valuable assistance of Dr Mitchell Steinberg and Kathleen Sweeney in the preparation of this manuscript.
Reprint requests to Robert J. Sweeney, PhD, Pacesetter Inc/St Jude Medical, One Lillihei Plaza, St Paul, MN 55117. E-mail email@example.com.
This work was supported in full by Lilly Research Laboratories Division of Eli Lilly and Co, of which the authors are full-time employees.
- Received April 8, 1996.
- Revision received June 25, 1996.
- Accepted July 8, 1996.
- Copyright © 1996 by American Heart Association
Sweeney RJ, Gill RM, Steinberg MI, Reid PR. Ventricular refractory period extension caused by defibrillation shocks. Circulation. 1990;82:965-972.
Sweeney RJ, Gill RM, Reid PR. Characterization of refractory period extension by transcardiac shock. Circulation. 1991;83:2057-2066.
Zhou X, Knisley SB, Wolf PD, Rollins DL, Smith WM, Ideker RE. Prolongation of repolarization time by electric field stimulation with monophasic and biphasic shocks in open chest dogs. Circ Res. 1991;68:1761-1767.
Zhou X, Knisley SB, Wolf PD, Rollins DL, Smith WM, Ideker RE. Action potential prolongation by biphasic and monophasic shocks during fibrillation. Pacing Clin Electrophysiol. 1991;14:689. Abstract.
Daubert JP, Frazier DW, Wolf PD, Franz MR, Smith WM, Ideker RE. Response of relatively refractory canine myocardium to monophasic and biphasic shocks. Circulation. 1991;84:2522-2538.
Dillon SM, Mehra R. Prolongation of ventricular refractoriness by defibrillation shocks may be due to additional depolarization of the action potential. J Cardiovasc Electrophysiol. 1992;3:442-457.
Gill RM, Sweeney RJ, Reid PR. Refractory period extension at fibrillatory activation rates. Pacing Clin Electrophysiol. In press.
Dillon SM. Synchronized repolarization after defibrillation shocks: a possible component of the defibrillation process demonstrated by optical recordings in rabbit heart. Circulation. 1992;85:1865-1878.
Knisley SB, Smith WM, Ideker RE. Effect of field stimulation on cellular repolarization in rabbit myocardium: implications for reentry induction. Circ Res. 1992;70:707-715.
Sweeney RJ, Gill RM, Reid PR. An in-vitro model of electrical defibrillation. J Am Coll Cardiol. 1993;21(suppl):307A. Abstract.
Swartz JF, Jones JL, Fletcher RD. Characterization of ventricular fibrillation based on monophasic action potential morphology in the human heart. Circulation. 1993;87:1907-1914.
Witkowski FX, Penkoske PA, Plonsey R. Mechanism of cardiac defibrillation in open-chest dogs with unipolar DC-coupled simultaneous activations and shock potential recordings. Circulation. 1990;82:244-260.
Chen PS, Shibata N, Dixon EG, Wolf PD, Danieley ND, Sweeney MB, Smith WM, Ideker RE. Activations during ventricular defibrillation in open-chest dogs: evidence of complete cessation and regeneration of ventricular fibrillation after unsuccessful shocks. J Clin Invest. 1986;77:810-823.
Shibata N, Chen PS, Dixon EG, Wolf PD, Danieley ND, Smith WM, Ideker RE. Epicardial activations after unsuccessful defibrillation shocks in dogs. Am J Physiol. 1988;255(Heart Circ Physiol 24):H902-H909.
Chen PS, Wolf PD, Melnick SD, Danieley ND, Smith WM, Ideker RE. Comparison of activations during ventricular fibrillation and following unsuccessful defibrillation shocks in open-chest dogs. Circ Res. 1990;66:1544-1560.
Zhou X, Daubert JP, Wolf PD, Smith WM, Ideker RE. Epicardial mapping of ventricular defibrillation with monophasic and biphasic shocks in dogs. Circ Res. 1993;72:145-160.
Sweeney RJ, Gill RM, Reid PR. Refractory period extension for monophasic versus biphasic shocks. J Am Coll Cardiol. 1994;23(special issue):422A. Abstract.
Sweeney RJ, Gill RM, Reid PR. Effect of intensity and timing on refractory period extension for like and opposite direction biphasic shocks. Am Heart J. 1992;124:835. Abstract.