Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Resuscitation. Author manuscript; available in PMC 2012 April 20.
Published in final edited form as:
PMCID: PMC3331677

Predictors of resuscitation outcome in a swine model of VF cardiac arrest: A comparison of VF duration, presence of acute myocardial infarction and VF waveform[star],[star][star]



Factors that affect resuscitation to a perfusing rhythm (ROSC) following ventricular fibrillation (VF) include untreated VF duration, acute myocardial infarction (AMI), and possibly factors reflected in the VF waveform. We hypothesized that resuscitation of VF to ROSC within 3 min is predicted by the VF waveform, independent of untreated VF duration or presence of acute MI.


AMI was induced by the occlusion of the left anterior descending coronary artery. VF was induced in normal (N = 30) and AMI swine (N = 30). Animals were resuscitated after untreated VF of brief (2 min) or prolonged (8 min) duration. VF waveform was analyzed before the first shock to compute the amplitude-spectral area (AMSA) and slope.


Unadjusted predictors of ROSC within 3 min included untreated VF duration (8 min vs 2 min; OR 0.11, 95%CI 0.02–0.54), AMI (AMI vs normal; OR 0.11, 95%CI 0.02–0.54), AMSA (highest to lowest tertile; OR 15.5, 95%CI 1.7–140), and slope (highest to lowest tertile; OR 12.7, 95%CI 1.4–114). On multivariate regression, untreated VF duration (P = 0.011) and AMI (P = 0.003) predicted ROSC within 3 min. Among secondary outcome variables, favorable neurological status at 24 h was only predicted by VF duration (OR 0.22, 95% CI 0.05–0.92).


In this swine model of VF, untreated VF duration and AMI were independent predictors of ROSC following VF cardiac arrest. AMSA and slope predicted ROSC when VF duration or the presence of AMI were unknown. Importantly, the initial treatment of choice for short duration VF is defibrillation regardless of VF waveform.

Keywords: Cardiopulmonary resuscitation, Myocardial infarction, Heart arrest, Ventricular fibrillation, Defibrillation, Ventricular fibrillation waveform

1. Introduction

Concepts of optimal timing for chest compressions and defibrillation following ventricular fibrillation (VF) have evolved substantially in the last several years. Optimal timing of chest compressions and defibrillation presumably relates to the underlying physiology of VF.1 In approximately the first 4 min of VF, the heart is in an electrical phase where defibrillation alone is likely to restore a perfusing rhythm. After the first few minutes, the heart transitions to a circulatory phase, where defibrillation alone rarely results in a perfusing rhythm (i.e., pre-shock or post-shock chest compressions are necessary).

Some investigators have proposed that the amplitude and frequency characteristics of the VF waveform, such as amplitude-spectral area (AMSA) and slope, can predict the need for a prompt defibrillation strategy vs a pre-shock chest compression strategy. However, it is unknown whether the VF waveform can provide any further predictive value if the untreated VF duration is already known, or how that prediction is affected by an AMI.

We hypothesized that successful and prompt resuscitation of VF to a perfusing rhythm can be predicted by AMSA and slope, independent of the duration of untreated VF and the presence of AMI. We tested this hypothesis in a swine model of 2 min vs 8 min of untreated VF cardiac arrest with or without AMI.

2. Methods

Experimental protocols were approved by the University of Arizona Institutional Animal Care and Use Committee and followed the guidelines of the American Physiologic Society. Protocols for the induction of anesthesia and instrumentation of domestic swine have been previously described.2

For control animals, VF was induced by a 100 Hz alternating current delivered to the right ventricle. In the AMI group, a steel plug was placed in the mid left anterior descending artery, proximal to the second diagonal branch, confirmed by fluoroscopy and contrast injection demonstrating complete coronary artery occlusion. If VF did not occur spontaneously by 15 min or if the pig developed severe hypotension, VF was induced. For the duration of untreated VF in both groups ventilation was discontinued and a continuous ECG was obtained via needle electrodes. Animals were randomized to 2 min or 8 min of untreated VF.

After the specified duration of untreated VF, resuscitation commenced with a defibrillation shock of 150 J of biphasic energy. If a perfusing rhythm was not immediately obtained after the shock (i.e., within 10 s), manual chest compressions were initiated at a metronome-guided rate of 100 compressions per minute to a depth of approximately one-third of the anterior–posterior diameter. Return of spontaneous circulation (ROSC) was defined as an unassisted pulse with a peak systolic aortic pressure greater than 50 mm Hg and pulse pressure at least 20 mm Hg for at least 1 min. After 2 min of chest compressions, a defibrillation shock (150 J) was given if the animal remained in VF. Resuscitation was continued until ROSC was obtained, or else efforts were terminated if ROSC was not achieved by 25 min. Epinephrine (0.02 mg/kg) was administered in the first minute of chest compressions, and repeated every 3 min until a perfusing rhythm was obtained for a total of three doses. The second scheduled dose of epinephrine did not occur until after 3 min of resuscitation. Swine that were successfully resuscitated were observed in the laboratory for 1 h, and returned to the observation pen for 24 h. Neurological status was evaluated by swine cerebral performance categories.3, 4 Category 1 was assigned to animals with normal levels of consciousness, gait, and feeding behavior, response to an approaching human and response to human restraint. Category 2 was assigned to animals with mild dysfunction in any of these areas, and category 3 was assigned to more severe dysfunction. Category 4 was assigned to animals in coma, and category 5 to dead animals. Categories 1 and 2 were regarded as a favorable neurological state. The primary outcome was ROSC within 3 min after the initiation of resuscitation, which was determined after initial pilot data. Secondary outcomes included 24-h survival, 24-h survival with favorable neurological state, and ROSC within 3 min after the initiation of resuscitation requiring only one shock.

Amplitude-spectral area (AMSA) and slope were blindly computed prior to the first shock from a 4.1-s interval encompassing 213 = 8192 data points. AMSA was blindly calculated as the summed product of frequency and square root of power at that frequency, from 4 to 20 Hz, from a Fast Fourier transform. Slope was blindly computed as the median of the absolute value of differences in signal voltage every 5 ms from the same 4.1-s interval prior to the shock.

The effects of untreated VF duration and AMI upon waveform characteristics were determined by ANOVA. An unadjusted logistic regression was performed to determine factors that predicted the outcome variables, and all significant factors on univariate analysis were further assessed by multivariate logistic regression. Odds ratios for waveform characteristics were assessed by tertiles.

3. Results

A total of 60 swine were studied, with N = 15 in each of four groups: normal coronary arteries with untreated VF for 2 min, normal coronary arteries with untreated VF for 8 min, AMI with untreated VF for 2 min, and AMI with untreated VF for 8 min duration. Baseline characteristics are presented in Table 1. Animals in the AMI group were somewhat heavier compared with the normal swine group (P = 0.003 by ANOVA).

Table 1
Baseline characteristics.

Mean (±SEM) values of AMSA and slope are shown in Table 2 according to VF duration and presence of AMI. AMSA was significantly dependent on untreated VF duration (P < 0.001) and borderline dependent on presence of AMI (P = 0.053). Slope was significantly dependent on both untreated VF duration (P < 0.001) and presence of acute MI (P = 0.031).

Table 2
Waveform characteristics by VF duration and presence of AMI.

For untreated VF of 2 min, 15 of 15 normal swine and 13 of 15 AMI swine achieved ROSC within 3 min (Table 3). For untreated VF of 8 min, 13 of 15 normal swine and 5 of 15 AMI swine achieved ROSC within 3 min. The untreated VF duration (2 min vs 8 min), presence of acute myocardial infarction, and highest tertiles of AMSA and slope were all significant univariate (unadjusted) predictors of achieving ROSC within 3 min of initiation of resuscitation (Table 4). In a multivariate regression, only untreated VF duration (P = 0.011) and presence of AMI (P = 0.003) remained significant predictors of achieving ROSC within 3 min.

Table 3
Outcome variables.
Table 4
Odds ratio and 95% confidence intervals for unadjusted (univariate) predictors of outcome variables.

Of the secondary outcome variables, favorable neurological outcome was also predicted by untreated VF duration (Table 4) (but not by the presence of AMI, AMSA or slope). In the 8 min of untreated VF groups, ROSC within 3 min of the initiation of resuscitation rarely occurred after only one shock (i.e., 2/15 normal swine and 0/15 AMI swine). By unadjusted analysis, untreated VF duration and the waveform parameters of AMSA and slope were each significant predictors of ROSC within 3 min with one shock (Table 4). Only untreated VF duration (P = 0.001) remained a significant predictor by multivariate regression analysis, although AMSA (P = 0.05) and slope (P = 0.05) were borderline significant predictors. Nearly all swine (29/30 normal swine and 24/30 AMI) survived 24 h (Table 3). Not surprisingly, none of the evaluated factors was a significant predictor of 24 h survival.

4. Discussion

This investigation has demonstrated that unadjusted (univariate) predictors of the achievement of ROSC within 3 min of the initiation of resuscitation included VF waveform characteristics of AMSA and slope, untreated VF duration (2 min vs 8 min), and presence of an AMI. This investigation supports the utility of the VF waveform to predict the achievement of ROSC, consistent with previous human and animal studies.58 However, this study further establishes that only untreated VF duration (2 min vs 8 min) and the presence of AMI were independent predictors for the attainment of ROSC within 3 min of the initiation of resuscitation. The waveform characteristics of AMSA and slope, which are affected by AMI and duration of untreated VF, did not improve the ability to predict the attainment of ROSC within 3 min if untreated VF duration and presence AMI were already known.

This study also provides further evidence that untreated short duration VF (e.g., witnessed VF in a public setting with an AED available, such as an airport or casino)9, 10 should be treated with prompt defibrillation, and that an analysis of the VF waveform will probably not improve the ability to predict ROSC. These data suggest that prompt defibrillation is also the treatment of choice for short duration untreated VF with an AMI regardless of the VF waveform. However, for an unwitnessed arrest without knowledge of myocardial milieu or untreated VF duration, these data imply that the VF waveform information may be helpful in deciding whether a shock-first or pre-shock chest compression strategy is more likely to result in successful initial resuscitation.

This study also has demonstrated that a favorable neurological outcome was only predicted by the duration of untreated VF (2 min vs 8 min), and not by the presence of AMI or VF waveform characteristics.

We also examined the likelihood of achieving ROSC within 3 min without requiring a second defibrillation shock. We found that for ROSC within 3 min, untreated VF duration was the critical determinant. Essentially no animals with prolonged untreated VF could be resuscitated without a second shock, and the VF waveform characteristics of AMSA and slope were only borderline predictive.

For prolonged VF, the concept of delaying defibrillation to provide pre-shock chest compressions was supported by an observational study from Seattle11 and a randomized controlled trial from Norway.12 However, this resuscitation strategy has been questioned, as two other randomized controlled human studies13, 14 did not show a benefit of pre-shock compressions for either short duration or prolonged duration VF. Perhaps other unmeasured factors, such as presence or absence of AMI, affected these outcomes. For example, our previous studies of prolonged untreated VF (8–10 min) in swine without AMI demonstrated that pre-shock chest compressions can be beneficial3, 4. In contrast, swine with AMI had much worse outcomes after pre-shock chest compressions following 8 min of untreated VF compared with a shock-first strategy.15 Importantly, autopsy studies have demonstrated an incidence of acute coronary thrombosis ranging from 15 to 80%.1618

Nonetheless, while VF duration and the presence of acute MI impact upon resuscitation, it is uncertain how to optimally carry out resuscitation when these factors are unknown. Therefore, there has been interest in identifying a surrogate for this information in the properties of the VF waveform itself or ideally a waveform property that is itself an independent predictor of ROSC.

Although duration of untreated VF and presence of AMI appear to be the most best predictors of resuscitation success, VF waveform characteristics are attractive additional sources of information because onset of VF and presence of AMI are often unknown. Various VF waveform characteristics, including AMSA and slope have been demonstrated to predict the ROSC in population and swine studies58. This swine investigation supports the predictive value of AMSA and slope when duration of untreated VF and presence of AMI are unknown.

4.1. Limitations

As with all animal cardiac arrest experiments, there are multiple potential experimental problems, including relevance to human prehospital VF cardiac arrest. We attempted to model elements commonly occurring with human VF, including both short duration untreated VF and longer duration untreated VF. In addition, we included animals with and without AMI. Although this experiment could not be blinded by its very nature, each animal was randomly assigned to one of the two treatment groups, the resuscitation and post-resuscitation protocols were standardized and strictly followed, and the VF waveforms were blindly analyzed. Moreover, outcomes and other end-points were clearly defined a priori to minimize investigator bias.

The optimal approach to model AMI in animal studies is unknown. In this study, AMI was induced by the placement of a steel plug beyond the first diagonal branch. In a prior study we placed the plug in the proximal left anterior descending artery.19 However, that resulted in a large MI accompanied by cardiogenic shock, and a high mortality in the first 24 h. In this study, as in previous studies,2, 2022 we chose to place the plug in a more distal location so that swine would be likely to survive 24 h to complete a neurological assessment. Presumably, these findings are most relevant to humans with moderate size acute myocardial infarctions. Furthermore, most patients with more proximal coronary obstructions are probably unlikely to survive.

5. Conclusion

In this swine model of VF, untreated VF duration and AMI were independent predictors of ROSC following VF cardiac arrest, whereas VF waveform analyses were not. Nevertheless, VF waveform parameters of AMSA and slope predicted ROSC when untreated VF duration or the presence of AMI was unknown.


We thank Duane Sherrill, PhD for his review of the statistical methods in this investigation.

Dr. Robert Berg had research grants from the NIH to study post-shock chest compressions in swine after prolonged VF and received research funding from Laerdal.


[star]A Spanish translated version of the abstract of this article appears as Appendix in the final online version at doi:10.1016/j.resuscitation.2009.08.023.

[star][star]This work was funded by a grant from the American Heart Association, AHA grant 0855587G.

Conflict of interest statements

There are no conflicts of interest to disclose for Julia Indik, Madhan Shanmugasundaram, Ronald Hilwig, Karl Kern, Mathias Zuercher, Daniel Allen or Amanda Valles.


1. Weisfeldt ML, Becker LB. Resuscitation after cardiac arrest: a 3-phase time-sensitive model. JAMA. 2002;288:3035–3038. [PubMed]
2. Indik JH, Donnerstein RL, Hilwig RW, et al. The influence of myocardial substrate upon ventricular fibrillation waveform: a swine model of acute and post-myocardial infarction. Crit Care Med. 2008;36:2136–2142. [PMC free article] [PubMed]
3. Berg RA, Hilwig RW, Kern KB, Ewy GA. Precountershock cardiopulmonary resuscitation improves ventricular fibrillation median frequency and myocardial readiness for successful defibrillation from prolonged ventricular fibrillation: a randomized, controlled swine study. Ann Emerg Med. 2002;40:563–570. [PubMed]
4. Berg RA, Hilwig RW, Ewy GA, Kern KB. Precountershock cardiopulmonary resuscitation improves initial response to defibrillation from prolonged ventricular fibrillation: a randomized, controlled swine study. Crit Care Med. 2004;32:1352–1357. [PubMed]
5. Young C, Bisera J, Gehman S, Snyder D, Tang W, Weil MH. Amplitude spectrum area: measuring the probability of successful defibrillation as applied to human data. Crit Care Med. 2004;32(Suppl):S356–S358. [PubMed]
6. Povoas HP, Weil MH, Tang W, Bisera J, Klouche K, Barbatsis A. Predicting the success of defibrillation by electrocardiographic analysis. Resuscitation. 2002;53:77–82. [PubMed]
7. Marn-Pernat A, Weil MH, Tang W, Pernat A, Bisera J. Optimizing timing of ventricular defibrillation. Crit Care Med. 2001;29:2360–2365. [PubMed]
8. Neurauter A, Eftestol T, Kramer-Johansen J, et al. Prediction of counter-shock success using single features from multiple ventricular fibrillation frequency bands and feature combinations using neural networks. Resuscitation. 2007;73:253–263. [PubMed]
9. Valenzuela T, Roe DJ, Nichol G, Clark LL, Spaite DW, Hardman RG. Outcomes of rapid defibrillation by security officers after cardiac arrest in casinos. New Engl J Med. 2000;343:1206–1209. [PubMed]
10. Caffrey SL, Willoughby PJ, Pepe PE, Becker LB. Public use of automated external defibrillators. N Engl J Med. 2002;347:1242–1247. [PubMed]
11. Cobb LA, Fahrenbruch CE, Walsh TR, et al. Influence of cardiopulmonary resuscitation prior to defibrillation in patients with out-of-hospital ventricular fibrillation. JAMA. 1999;281:1182–1188. [PubMed]
12. Wik L, Hanse TB, Fylling F, et al. Delaying defibrillation to give basic cardiopulmonary resuscitation to patients with out-of-hospital ventricular fibrillation. A randomized trial. JAMA. 2003;289:1389–1395. [PubMed]
13. Jacobs IG, Finn JC, Oxer HF, Jelinek GA. CPR before defibrillation in out-of-hospital cardiac arrest: a randomized trial. Emerg Med Aust. 2005;17:39–45. [PubMed]
14. Baker PW, Conway J, Cotton C, et al. Defibrillation or cardiopulmonary resuscitation first for patients with out-of-hospital cardiac arrests found by paramedics to be in ventricular fibrillation? A randomised control trial. Resuscitation. 2008;79:424–431. [PubMed]
15. Indik JH, Hilwig RW, Zuercher M, Kern KB, Berg MD, Berg RA. Pre-shock CPR worsens outcome from circulatory phase VF with acute coronary artery obstruction in swine. Circ: Arrhythmia Electrophysiol. 2009;2:179–184. [PMC free article] [PubMed]
16. Schmermund A, Schwartz RS, Adamzik M, et al. Coronary atherosclerosis in unheralded sudden coronary death under age 50: histo-pathologic comparison with ‘healthy’ subjects dying out of hospital. Atherosclerosis. 2001;155:499–508. [PubMed]
17. Engdahl J, Holmberg M, Karlson BW, Luepker R, Herlitz J. The epidemiology of out-of-hospital ‘sudden’ cardiac arrest. Resuscitation. 2002;52:235–245. [PubMed]
18. Fornes P, Lecomte D, Nicolas G. Sudden out-of-hospital coronary death in patients with no previous cardiac history. An analysis of 221 patients studied at autopsy. J Forensic Sci. 1993;38:1084–1091. [PubMed]
19. Indik JH, Donnerstein R, Berg R, Hilwig R, Berg M, Kern K. Ventricular fibrillation frequency characteristics are altered in acute myocardial infarction. Crit Care Med. 2007;35:1133–1138. [PubMed]
20. Kern KB, Ewy GA. Minimal coronary stenoses and left ventricular blood flow during CPR. Ann Emerg Med. 1992;21:1066–1072. [PubMed]
21. Kern KB, Lancaster L, Goldman S, Ewy GA. The effect of coronary artery lesions on the relationship between coronary perfusion pressure and myocardial blood flow during cardiopulmonary resuscitation in pigs. Am Heart J. 1990;120:324–333. [PubMed]
22. Berg RA, Kern KB, Hilwig RW, Ewy GA. Assisted ventilation during “bystander” CPR in a swine acute myocardial infarction model does not improve outcome. Circulation. 1997;96:4364–4371. [PubMed]