An up-down Bayesian, defibrillation efficacy estimator

Response-adaptive clinical trials: case studies in the medical literature

The past 15 years has seen many pharmaceutical sponsors consider and implement adaptive designs (AD) across all phases of drug development. Given their arrival at the turn of the millennium, we might think that they are a recent invention. That is not the case. The earliest idea of an AD predates Bradford Hill's MRC tuberculosis study, appearing in Biometrika in 1933. In this paper, we trace the development of response-ADs, designs in which the allocation to intervention arms depends on the responses of subjects already treated. We describe some statistical details underlying the designs, but our main focus is to describe and comment on ADs from the medical research literature. Copyright © 2016 John Wiley & Sons, Ltd.

A Grouped Up-and-Down Method Used for Efficacy Comparison Between Two Different Defibrillation Waveforms

Electrical defibrillation, which consists of delivering a therapeutic dose of the electrical current to the fibrillating heart with the aid of a defibrillator, is still the only effective way to treat life-threatening ventricular fibrillation (VF). However, the efficacy of electrical therapy for terminating VF is highly dependent on the waveform applied. When new defibrillation waveforms or techniques are developed, their efficacy needs to be accurately evaluated and compared to those in use. A common method for the comparison of defibrillation efficacy is to estimate and compare the individual defibrillation threshold (DFT) by constructing dose response curves or using an up-and-down method. Since DFT is calculated by repetitive and sequential shocks, there will be variability for each measurement and for each individual. This creates a considerable uncertainty for paired comparison. In this paper, a novel grouped up-and-down method is developed for the comparison of defibrillation efficacy between two different defibrillation waveforms or techniques. The efficacy of two commonly used biphasic defibrillation waveforms was compared in a porcine model of cardiac arrest using the developed method. Experimental results demonstrate that the proposed method is more sensitive for efficacy comparison and requires less defibrillation attempts compared with traditional DFT methods.

Simultaneous comparison of many triphasic defibrillation waveforms

Biphasic defibrillation waveforms are now accepted as being more effective at terminating ventricular fibrillation (VF) than monophasic waveforms. If two phases are better than one, this naturally leads to the hypothesis that additional phases improve efficacy. This study tests the hypothesis by adding one additional phase. We examined the efficacy of 18 different triphasic waveforms simultaneously.

We tested the rate of recovery, i.e., successful defibrillation, of 21 guinea pigs (820-1,050 g) using triphasic, monophasic and biphasic defibrillation waveforms. The biphasic and monophasic were control waveforms. VF was electrically induced twenty times per animal and a single defibrillation attempt was made using a test waveform VF episode. Every waveform was adjusted to the energy required to defibrillate that animal 50% of the time, using a biphasic waveform as a control. The success rate of each triphasic waveform was pair-wise compared to the biphasic and monophasic control using the adjusted McNemar statistical test.

Of the 18 triphasic waveforms tested, two were significantly poorer than the monophasic control (p<0.05). One was superior to the biphasic waveform (p<0.1), but not statistically so. We concluded that, while adding a phase to a monophasic waveform does improve efficacy, adding an additional phase to a biphasic waveform does not necessarily improve efficacy.

Cilostazol attenuates ventricular arrhythmia induction and improves defibrillation efficacy in swine

Previous reports demonstrated that cilostazol, a phosphodiesterase 3 inhibitor, affected cellular electrophysiology and reduced episodes of ventricular fibrillation (VF) in patients with Brugada syndrome. However, its effects on VF induction and defibrillation efficacy have never been investigated. We tested the hypothesis that cilostazol increases the VF threshold (VFT) and decreases the upper limit of vulnerability (ULV) and the defibrillation threshold (DFT). A total of 48 pigs were randomly assigned to defibrillation and VF induction studies. The diastolic pacing threshold (DPT), VFT, ULV, DFT, and effective refractory period were determined before and after the infusion of cilostazol at 6 mg/kg, 3 mg/kg, or vehicle. The DPT was significantly increased after administration of 3 and 6 mg/kg cilostazol. The ULV and DFT were significantly decreased after administration of 6 mg/kg cilostazol only. The ULV in the 6 mg/kg group had 12% lower peak voltage and 25% lower total energy, and the DFT had 13% lower peak voltage and 25% lower total energy. The VFT was not altered in any experimental group. This study shows that cilostazol administration significantly increased the DPT, which was associated with significantly reduced DFT and ULV.

Experimental evidence of improved transthoracic defibrillation with electroporation-enhancing pulses

There is considerable work on defibrillation wave form optimization. This paper determines the impedance changes during defibrillation, then uses that information to derive the optimum defibrillation wave form.

METHODS PART I

Twelve guinea pigs and six swine were used to measure the current wave form for square voltage pulses of a strength which would defibrillate about 50% of the time. In guinea pigs, electrodes were placed thoracically, abdominally and subcutaneously using two electrode materials (zinc and steel) and two electrode pastes (Core-gel and metallic paste).

RESULTS PART I

The measured current wave form indicated an exponentially increasing conductance over the first 3 ms, consistent with enhanced electroporation or another mechanism of time-dependent conductance. We fit this current with a parallel conductance composed of a time-independent component (g0 = 1.22 +/- 0.28 mS) and a time-dependent component described by g delta (1-e(-t/tau)), where g delta = 0.95 +/- 0.20 mS and tau = 0.82 +/- 0.17 ms in guinea pigs using zinc and Cor-gel. Different electrode placements and materials had no significant effect on this fit. From our fit, we determined the stimulating wave form that would theoretically charge the myocardial membrane to a given threshold using the least energy from the defibrillator. The solution was a very short, high voltage pulse followed immediately by a truncated ascending exponential tail.

METHODS PART II

The optimized wave forms and similar nonoptimized wave forms were tested for efficacy in 25 additional guinea pigs and six additional swine using methods similar to Part I.

RESULTS PART II

Optimized wave forms were significantly more efficacious than similar nonoptimized wave forms. In swine, a wave form with the short pulse was 41% effective while the same wave form without the short pulse was 8.3% effective (p < 0.03) despite there being only a small difference in energy (111 J versus 116 CONCLUSIONS: We conclude that a short pulse preceding a defibrillation pulse significantly improves efficacy, perhaps by enhancing electroporation.

Analysis of the defibrillation efficacy for 5-ms waveforms

INTRODUCTION

Empirical studies have shown that biphasic defibrillation waveforms are more efficacious than monophasic waveforms. However, a more systematic approach to waveform development might be more productive. This study tested 147 multiphasic waveforms uniformly sampled from all possible 5-ms waveforms.

METHODS AND RESULTS

One hundred ninety-eight guinea pigs (850-1,050 g) received 30 episodes of ventricular fibrillation followed by transthoracic defibrillation. The first 10 shocks were used to determine the ED(50) for a biphasic control. Then, 20 waveforms including 2 controls were tested once at the ED(50). Of the 147 waveforms tested here, 21 waveforms showed equivalent or greater efficacies than the biphasic control, with one being statistically more efficacious (P < 0.05). Two fundamental assumptions were addressed: (1) similarly efficacious waveforms are analytically similar, and (2) a single optimal waveform can be described. The mean percentage of similarly efficacious waveforms with similar shapes was greater than zero in the most efficacious 21 waveforms (P = 0.023), but less efficacious waveforms showed randomly distributed shapes. Cluster analysis revealed that the best waveforms share a major phase containing most of the defibrillation energy. The optimal waveform shape extrapolated from the sample waveforms was a 2.5/1-ms biphasic-type waveform (highest correlation r = 0.701, P < 0.001).

CONCLUSION

This work supports the assumption that efficacious waveforms are similarly shaped and the notion that one single optimum exists.

Minimum energy single-shock internal atrial defibrillation in sheep

Well-tolerated internal atrial defibrillation shocks must be below the pain threshold, which has been estimated to be less than 1 Joule. Defibrillation of the atria with low energy is made possible by delivering shocks at the low end of the defibrillation dose-response curve. We studied low-energy defibrillation in sheep to test the hypothesis that the energy that defibrillates the atria 10% of the time (ED10) is less than 1 Joule. The ED10 was estimated in seven sheep with rapid pacing induced chronic atrial fibrillation (AF). Low-energy defibrillation shocks were delivered from coronary sinus (CS) to superior vena cava (SVC) and the ED10 and ED50 (energy that defibrillates the atria 50% of the time) were then calculated using logistic regression. The mean ratio of ED10 to ED50 was 0.50, indicating that on average, the ED10 was equal to half of the ED50. ED10 shocks had energies ranging from 1.2 to 5.8 Joules. These results suggest that painless single-shock low-energy defibrillation may not be feasible.

The Strength-Duration Relationship of Monophasic Waveforms with Varying Capacitance Sizes in External Defibrillation

The shape of the shock waveform influences defibrillation efficacy. However, the optimal combination between capacitance size and truncation/tilt which can determine monophasic waveform's shape, has not been determined for external defibrillation. The purpose of this study was to assess the effects of varying capacitance and tilt on external defibrillation using exponential monophasic waveforms. In a pig model of external defibrillation (n = 10, 30 +/- 6 kg), nine exponential monophasic waveforms combining three capacitance values (30 microF, 60 microF, and 120 microF) and three tilt values (55%, 75%, and 95%) were tested randomly. The energy and leading edge voltage at 50% defibrillation success (E50 and V50) were used to evaluate defibrillation efficacy. E50 and V50 were determined by the Bayesian technique. The lowest stored E50 for the 30microF, 60 microF, and 120 microF waveforms were 90 +/- 12 J (95% tilt), 106 +/- 45 J (55% tilt), and 107 +/- 52 J (75% tilt), respectively. The lowest V50 for the 30 microF, 60 microF, and 120 microF waveforms were 2,439 +/- 166 V (95% tilt), 1,849 +/- 375 V (55% tilt), and 1,301 +/- 322 V (75% tilt), respectively. The average current at external defibrillation threshold demonstrated a strength versus pulse duration relationship similar to that seen with pacing. Reducing capacitance has the same effect as truncating the waveform. The E50 is more sensitive to tilt values changes in larger capacitance waveforms. This study suggests that the optimal combination between capacitance and tilt may be 120 microF and 55%-75% for external defibrillation.

Left ventricular geometry immediately following defibrillation: shock-induced relaxation

A previous two-dimensional (2D) ultrasound study suggested that there is relaxation of the myocardium after defibrillation. The 2D study could not measure activity occurring within the first 33 ms after the shock, a period that may be critical for discriminating between shock- and excitation-induced relaxation. The objective of our study was to determine the left ventricular (LV) geometry during the first 33 ms after defibrillation. Biphasic defibrillation shocks were delivered 5-50 s after the induction of ventricular fibrillation in each of the seven dogs. One-dimensional, short-axis ultrasound images of the LV cavity were acquired at a rate of 250 samples/s. The LV cavity diameter was computed from 32 ms before to 32 ms after the shock. Preshock and postshock percent changes in LV diameter were analyzed as a function of time with the use of regression analysis. The normalized mean pre- and postshock slopes (0.2 +/- 2.2 and 3.3 +/- 7.9% per 10 ms) were significantly different (P < 0.01). The postshock slope was positive (P < 0.005). Our results confirm that the bulk of the myocardium is relaxing immediately after defibrillation.

Defibrillation and the geometry of the heart: a novel measurement with implications for defibrillation mechanisms

We present a novel measurement for studying defibrillation mechanisms: the time course of changes in the size of the left ventricular (LV) cavity within 500 ms following defibrillation. Mechanical changes can be linked to electrical mechanisms via an understanding of excitation-contraction coupling. Eight mongrel dogs were internally defibrillated 5-50 seconds (including backup shocks) after the onset of 20 ventricular fibrillation (VF) episodes per animal. Two dimensional, short axis, LV cavity, ultrasound images were recorded at 30 frames per second just prior to inducing VF, during defibrillation and following the shock. Each frame was individually analysed to yield the LV cavity area as a function of time. Defibrillation shocks were followed by a highly reproducible phenomenon: (1) a dramatic and rapid increase in LV area, (2) a more or less prominent LV area plateau and (3) a decrease in the LV area. The peak fractional area increase ranged from 1.65 to 4.64 times larger than the baseline (LV area just prior to defibrillation), averaging 2.18 +/- 0.686. Successful shocks took significantly longer (p < 0.01) to return to 1.3 times the baseline (407 +/- 209 ms) than unsuccessful shocks (296 +/- 130 ms). Extrapolating to electrical mechanisms, our novel measurement demonstrates that defibrillation causes immediate relaxation and therefore suggests a significant role for deexcitation in defibrillation.

New approach to biphasic waveforms for internal defibrillation: fully discharging capacitors

INTRODUCTION

The use of two independent, fully discharging capacitors for each phase of a biphasic defibrillation waveform may lead to the design of a simpler, smaller, internal defibrillator. The goal of this study was to determine the optimal combination of capacitor sizes for such a waveform.

METHODS AND RESULTS

Eight full-discharge (95/95% tilt), biphasic waveforms produced by several combinations of phase-1 capacitors (30, 60, and 90 microF) and phase-2 capacitors (1/3, 2/3, and 1.0 times the phase-1 capacitor) were tested and compared to a single-capacitor waveform (120 microF, 65/65% tilt) in a pig ventricular fibrillation model (n = 12, 23+/-2 kg). In the full-discharge waveforms, phase-2 peak voltage was equal to phase-1 peak voltage. Shocks were delivered between a right ventricular lead and a left pectoral can electrode. E50s and V50s were determined using a ten-step Bayesian process. Full-discharge waveforms with phase-2 capacitors of < or =40 microF had the same E50 (6.7+/-1.7 J to 7.3+/-3.9 J) as the single-capacitor truncated waveform (7.3+/-3.7 J), whereas waveforms with phase-2 capacitors of > or =60 microF had an extremely high E50 (14.5+/-10.8 J or greater, P < 0.05). Moreover, of the former set of energy-efficient waveforms, those with phase-1 capacitors of > or =60 microF additionally exhibited V50s that were equivalent to the V50 of the single-capacitor waveform (344+/-65 V to 407+/-50 V vs 339+/-83 V).

CONCLUSION

Defibrillation efficacy can be maintained in a full-discharge, two-capacitor waveform with the proper choice of capacitors.

Hemodynamic collapse, geometry, and the rapidly paced upper limit of ventricular vulnerability to fibrillation by T-wave stimulation

There is an upper limit to the vulnerability (ULV) of the ventricles to fibrillation (VF) induced by T-wave stimuli. Across species, disease states, and pharmacological treatments, the ULV is correlated to the defibrillation threshold (DF50). However, one factor known to increase the ULV far above the DF50 is rapid pacing. In this article we test the hypothesis that this increase is owing to an accompanying hemodynamic collapse or geometric change. In 18 dogs, T-wave stimuli were delivered from transvenous defibrillating electrodes. The T-wave shock strength that induced VF 50% of the time (the ULV50) was measured using a 10-step Bayesian up-down protocol. T-wave stimuli were delivered after 15 paced beats at one of several rates: normal (80% of the R-R interval), rapid (the interval just fast enough to cause hemodynamic collapse), or 10 milliseconds greater than rapid (which did not cause hypotension). We measured the geometry of the left ventricle at the moment of T-wave stimulation using linear ultrasound. Rapid pacing significantly increased the ULV50 above the normal rate ULV (507 +/- 62.9 vs 379 +/- 70.6 V, P < .005, n = 18), even in the subset without hemodynamic collapse (505 +/- 84.4 vs 394 +/- 66.5 V, P < .005, n = 6). No significant geometric changes were noted between rapid (19.8 mm) and normal (20.6 mm, n = 6, P < NS) pacing, but QT interval reduction appears to correlate with the ULV50 (QT vs ULV50, r > 0, P < .01). Rapid pacing can dramatically increase the measured ULV50. The most likely cause is a concurrent change in the electrophysiology, eg, QT or APD, of the myocardium. As the only known factor to consistently alter the relationship between ULV and the DF50, rapid pacing offers a unique opportunity for the study of the link between defibrillation and ULV testing.

Experimental cardiac tachyarrhythmias in guinea pigs

Despite years of intense research into the mechanisms of defibrillation, there remain many unanswered questions. In many fields, hypotheses are first tested in rodent models before confirming the results in larger animals. This work suggests the guinea pig as a rodent model for defibrillation. Twenty-eight guinea pigs were studied, all male retired breeders weighing over 900 g. T-wave stimuli (upper limit of vulnerability [ULV]) were given after 15 rapid pacing beats, since the rapid pacing has been suggested to extend the tachyarrhythmia. Defibrillation (DF) was attempted after 5 seconds. The correlation between the ULV50 and DF50 in guinea pigs (0.82, n = 8) is very close to that seen in dogs (0.85). Also, the sensitivity of the DF50 to waveform is similar (476 +/- 176 for monophasic vs 364 +/- 94 V for biphasic P < 0.005, n = 10). The dose-response curve widths (2.3 +/- 1.7 for ULV vs 1.9 +/- 1.8 for defibrillation, n = 10) show the same trend of increasing curve widths for ULV, and similar magnitude to dogs (mean 1.8). We rarely (<1.5%) observed spontaneous conversion in less than 10 seconds. The guinea pig can be used as a model for defibrillation as it shows many of the same characteristics as dogs.

External exponential biphasic versus monophasic shock waveform: efficacy in ventricular fibrillation of longer duration

Ventricular fibrillation (VF) duration may be a factor in determining the defibrillation energy for successful defibrillation. Exponential biphasic waveforms have been shown to defibrillate with less energy than do monophasic waveforms when used for external defibrillation. However, it is unknown whether this advantage persists with longer VF duration. We tested the hypothesis that exponential biphasic waveforms have lower defibrillation energy as compared to exponential monophasic waveforms even with longer VF duration up to 1 minute. In a swine model of external defibrillation (n = 12, 35 +/- 6 kg), we determined the stored energy at 50% defibrillation success (E50) after both 10 seconds and 1 minute of VF duration. A single exponential monophasic (M) and two exponential biphasic (B1 and B2) waveforms were tested with the following characteristics: M (60 microF, 70% tilt), B1 (60/60 microF, 70% tilt/3 ms pulse width), and B2 (60/20 microF, 70% tilt/3 ms pulse width) where the ratio of the phase 2 leading edge voltage to that of phase 1 was 0.5 for B1 and 1.0 for B2. E50 was measured by a Bayesian technique with a total often defibrillation shocks in each waveform and VF duration randomly. The E50 (J) for M, B1, and B2 were 131 +/- 41, 57 +/- 18,* and 60 +/- 26* with 10 seconds of VF duration, respectively, and 114 +/- 62, 77 +/- 45,* and 72 +/- 53* with 1 minute of VF duration, respectively (*P < 0.05 vs M). There was no significant difference in the E50 between 10 seconds and 1 minute of VF durations for each waveform. We conclude that (1) the E50 does not significantly increase with lengthening VF durations up to 1 minute regardless of the shock waveform, and (2) external exponential biphasic shocks are more effective than monophasic waveforms even with longer VF durations.

Fully discharging phases. A new approach to biphasic waveforms for external defibrillation

BACKGROUND

Phase-2 voltage and maximum pulse width are dependent on phase-1 pulse characteristics in a single-capacitor biphasic waveform. The use of 2 separate output capacitors avoids these limitations and may allow waveforms with lower defibrillation thresholds. A previous report also suggested that the optimal tilt may be >70%. This study was designed to determine an optimal biphasic waveform by use of a combination of 2 separate and fully (95% tilt) discharging capacitors.

METHODS AND RESULTS

We performed 2 external defibrillation studies in a pig ventricular fibrillation model. In group 1, 9 waveforms from a combination of 3 phase-1 capacitor values (30, 60, and 120 microF) and 3 phase-2 capacitor values (0=monophasic, 1/3, and 1.0 times the phase-1 capacitor) were tested. Biphasic waveforms with phase-2 capacitors of 1/3 times that of phase 1 provided the highest defibrillation efficacy (stored energy and voltage) compared with corresponding monophasic and biphasic waveforms with the same capacitors in both phases except for waveforms with a 30-microF phase-1 capacitor. In group 2, 10 biphasic waveforms from a combination of 2 phase-1 capacitor values (30 and 60 microF) and 5 phase-2 capacitor values (10, 20, 30, 40, and 50 microF) were tested. In this range, phase-2 capacitor size was more critical for the 30-microF phase-1 than for the 60-microF phase-1 capacitor. The optimal combinations of fully discharging capacitors for defibrillation were 60/20 and 60/30 microF. Conclusions-Phase-2 capacitor size plays an important role in reducing defibrillation energy in biphasic waveforms when 2 separate and fully discharging capacitors are used.

A four-shock Bayesian up-down estimator of the 80% effective defibrillation dose

INTRODUCTION

New defibrillation techniques are often compared to standard approaches using the defibrillation threshold. However, inference from thresholding data necessitates extrapolation from reactions to relatively ineffective shocks, an error prone procedure requiring large sample sizes for hypothesis testing and large safety margins for defibrillator implantation. In contrast, this article presents a clinically validated statistical model of a minimum error, four-shock defibrillation testing protocol for estimating the 80% effective defibrillation strength for a given patient (ED80).

METHODS AND RESULTS

A Bayesian statistical model was constructed assuming that the defibrillation dose-response curve is sigmoidal, and the ED80 is between 150 and 750 V. The model was used to design a minimum predicted error testing protocol and estimates. To prospectively validate the testing protocol and estimates, 170 patients received voltage-programmed biphasic testing. Four fibrillation episodes were induced and terminated in each patient according to the Bayesian up-down protocol. In addition, a validation attempt was made at the estimated ED80 rounded up to the nearest 50 V. In order to estimate the safety margin, in 136 patients, a defibrillation attempt was made at the rounded ED80 + 100 V. Of the 170 attempts at the rounded ED80, 143 (84%) attempts terminated fibrillation. Of the 136 attempts at the rounded ED80 + 100 V, 133 (98%) were effective.

CONCLUSIONS

The four-shock Bayesian up-down protocol is the first clinical protocol to accurately predict an ED80 voltage. A 100 V increment above the ED80 provides an adequate safety margin. This simple and accurate method for estimating a highly effective defibrillation dose may be a valuable tool for population-based clinical hypothesis testing, as well as defibrillator implantation.

Energy levels for defibrillation: what is of real clinical importance?

Today, transthoracic and intracardiac defibrillation offer a well-accepted and widely used form of therapy for patients with life-threatening ventricular arrhythmias. Despite the wide clinical use of defibrillators, the mechanisms by which an electrical shock halts fibrillation are still not completely understood. During a shock, different amounts of current flow through the different parts of the heart and the current distribution is highly uneven. This current distribution is affected by changes in the shock potential gradient through the heart, changes in fiber orientation, and changes in myocardial conductivity caused by connective tissue barriers. It would be ideal if the potential gradient distribution throughout the ventricles could be measured directly for each individual patient during defibrillator implantation and follow-up and the shock strength could be programmed based on this measurement, but so far this is not possible. A more feasible approach is to determine, by trial and error, the magnitude of the shock strength delivered through the defibrillation electrodes for successful defibrillation. There is no distinct threshold value above which all shocks succeed and below which all shocks fail to defibrillate. Rather, increasing shock strength increases the likelihood the shock will succeed. Therefore, instead of a distinct defibrillation threshold, a probability of success curve exists. However, increasing the shock strength above an optimal range can actually decrease the success rate for defibrillation. One possible explanation is that the high voltage gradients caused by such large shocks damage cells and result in postshock arrhythmias that may reinitiate fibrillation. Another problem that can affect the probability of defibrillation success for a particular programmed energy setting is that the shock strength required for defibrillation may increase over time due to (1) the growth of fibrotic tissue around the defibrillation electrode; (2) migration of the lead; (3) acute ischemia; or (4) other changes in the underlying cardiac disease (e.g., worsening of heart failure). Such possible increases in the defibrillation shock strength requirement should be compensated for before they occur by adding a margin of safety to the shock strength needed for effective defibrillation. When programming an implantable defibrillator, it is important to keep in mind that the defibrillation shock should be (1) strong enough to defibrillate at least 98% of the time with the first shock; (2) weak enough not to cause severe post-shock arrhythmias or reinitiation of fibrillation; but (3) strong enough to compensate for changes of defibrillation energy requirements over time. This usually can be accomplished by setting the defibrillator 7-10 J higher than the defibrillation threshold determined by a standard step-down protocol.

Optimal small-capacitor biphasic waveform for external defibrillation: influence of phase-1 tilt and phase-2 voltage

BACKGROUND

Biphasic waveforms have been reported to be more efficacious than monophasic waveforms for external defibrillation. This study examined the optimal phase-1 tilts and phase-2 leading-edge voltages with small capacitors (60 and 20 microF) for external defibrillation. We also assessed the ability of the "charge-burping" model to predict the optimal waveforms.

METHODS AND RESULTS

Two groups of studies were performed. In group 1, 9 biphasic waveforms from a combination of 3 phase-1 tilt values (30%, 50%, and 70%) and 3 phase-2 leading-edge voltage values (0.5, 1.0, and 1.5 times the phase-1 leading-edge voltage, V1) were tested. Phase-2 pulse width was held constant at 3 ms in all waveforms. Two separate 60- microF capacitors were used in each phase. The energy value that would produce a 50% likelihood of successful defibrillation (E50) decreased with increasing phase-1 tilt and increased with increasing phase-2 leading-edge voltage except for the 30% phase-1 tilt waveforms. In group 2, 9 waveforms were identical to the waveforms in group 1, except for a 20- microF capacitor for phase 2. E50 decreased with increasing phase-1 tilt. Phase-2 leading-edge voltage of 1.0 to 1.5 V1 appeared to minimize E50 for phase-1 tilt of 50% and 70% but worsened E50 for phase-1 tilt of 30%. There was a significant correlation between E50 and residual membrane voltage at the end of phase 2, as calculated by the charge-burping model in both groups (group 1, R2=0.47, P<0.001; group 2, R2=0.42, P<0.001).

CONCLUSIONS

The waveforms with 70% phase-1 tilt were more efficacious than those with 30% and 50%. The relationship of phase-2 leading-edge voltage to defibrillation efficacy depended on phase-2 capacitance. The charge-burping model predicted the optimal external biphasic waveform.

The effect of inducing ventricular fibrillation with 50-Hz pacing versus T wave stimulation on the ability to defibrillate

When testing an ICD, there are at least two techniques for inducing ventricular fibrillation: (1) high frequency (approximately equal to 50 Hz) pacing; and (2) a single T wave stimulus. It is generally assumed that these two methods yield similar results. This study directly tested this assumption. In six dogs, one defibrillation electrode was placed in the right ventricular (RV) apex and the second was placed cutaneously on the left thorax. All defibrillation and T wave stimuli were biphasic between these two electrodes. Pacing was monophasic from the tip of the RV catheter to the cutaneous patch. The voltage which defibrillates 50% of the time (DF50) was measured using a 10-step Bayesian up-down method. Observations for two DF50 measurements were randomly interleaved. For one DF50 measurement, fibrillation was induced with 99 pacing stimuli at a 20-ms pacing interval (50-Hz pacing). For the second DF50 measurement, fibrillation was induced with a single defibrillation shock of approximately 1/2 J delivered at a time corresponding to the peak of the T wave in the lead II electrogram (T wave stimuli). The average DF50 when measured after fibrillation induced with 50-Hz pacing was 379 +/- 54.6 V, as compared to 382 +/- 50.3 V when fibrillation was induced with T wave stimuli. The difference of 3 V was not statistically significant. If these results are confirmed in humans, it is reasonable to assume that the efficacy of a defibrillation shock is the same whether T wave stimuli or 50-Hz pacing are used to induce fibrillation.

The ventricular defibrillation and upper limit of vulnerability dose-response curves

INTRODUCTION

A stimulus delivered in the T wave of a paced cardiac cycle can induce ventricular fibrillation (VF). If the stimulus strength is increased, the probability of inducing VF decreases. This study determines an ideal mathematical model (a dose-response curve) for the relationship between the shock strength and the probability of inducing VF or defibrillating.

METHODS AND RESULTS

Defibrillating electrodes were implanted in the right ventricle and superior vena cava in 16 pigs. The electrode in the vena cava was electrically connected to a cutaneous patch. The same electrodes were used for both VF induction and defibrillation. T wave stimuli were given at the peak of the T wave according to a modified up-down protocol (40 V up, 20 V down). When a T wave stimulus induced VF, a defibrillation stimulus was delivered 10 seconds later, also according to the modified up-down protocol. Exponential, logistic, log-dose logistic, piecewise linear and Box-Tiao dose-response curves were fit to the resulting data using the maximum likelihood method. For the defibrillation data, it was found that only the logistic and Box-Tiao curves fit all of the animals (P < 0.05). For VF induction, only the Box-Tiao curve fit all of the animals (P < 0.05). Extrapolating along a dose-response curve that did not fit to a shock strength with a very low probability of inducing VF or a very high probability of defibrillating yielded errors as great as 610 V.

CONCLUSION

The Box-Tiao dose-response curve is the best single choice for fitting VF induction or defibrillation datasets.