The ventricular fibrillation waveform in relation to shock success in early vs. late phases of out-of-hospital cardiac arrest
Jos Thannhauser M.Sc.1, Joris Nas M.D.1, Gjerrit Meinsma Ph.D.2, Hans J. Zwart Ph.D. 2, Pierre van Grunsven M.D. Ph.D. 3, Menko-Jan de Boer M.D. Ph.D.1, Niels Van Royen M.D. Ph.D. 1, Judith L. Bonnes M.D. Ph.D. 1, Marc A. Brouwer M.D. Ph.D.1
1 Department of Cardiology, Radboud University Medical Center, Nijmegen, The Netherlands Address: Geert Grooteplein Zuid 10, 6525 GA Nijmegen, The Netherlands
2 Faculty of Electrical Engineering, Mathematics and Computer Sciences, Department of Applied Mathematics, University of Twente, Enschede, The Netherlands
Address: P.O. Box 217, 7500 AE Enschede, The Netherlands
3 Regional Ambulance Service Gelderland-Zuid, Nijmegen, The Netherlands Address: Professor Bellefroidstraat 11, 6525 AG Nijmegen, The Netherlands
Running title: AMSA in early vs. late phase of resuscitation Word count: abstract: 250; paper: 3003
Number of Tables: 3 Number of Figures: 1
Number of Supplementary Tables/Figures/Text: 1/1/1 Funding: None
Conflicts of interest: Prof. De Boer and Prof. Van Royen have conflicts of interest to declare, see page 14.
Address for correspondence: Jos Thannhauser, Radboud University Medical Center, Department of Cardiology 616, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands. Tel: +31 243616785,
Abstract
Background: The amplitude spectrum area (AMSA) of the ventricular fibrillation (VF) waveform predicts shock success as well as clinical outcome after out-of-hospital cardiac arrest (OHCA).
Recently, also AMSA-changes demonstrated prognostic value. As of yet, most studies focused on early shocks, while many patients require prolonged resuscitations. We studied AMSA and its changes in relation to shock success both in the early and the late phase of resuscitation.
Methods: Per-shock analysis on VF-episodes from a prospective OHCA-cohort (Nijmegen, The Netherlands). AMSA and AMSA-changes between shocks (ΔAMSA) were calculated from three-second VF-recordings prior to shock delivery. Shocks were categorised as early (1-3) or late (4-8). Shock success was defined as return of organised rhythm within one minute after the shock.
Results: Of all shocks, 48% was successful (216/448), without difference between early
(131/286=46%) and late shocks (85/162=52%), p=0.18. Early shock success markedly varied from 20% to 70% with increasing quartiles of AMSA (p-for-trend<0.001). For late shocks, we also observed an association with AMSA (p-for-trend=0.04), and shock success ranged from 40% to 65%. Higher values of ΔAMSA were significantly associated with shock success in the early, but not in the late phase of resuscitation.
Conclusion: AMSA relates to shock success during the entire resuscitation, but associations were most markedly present for early shocks. AMSA-changes were also associated with shock success, but only in the early phase of resuscitation. We implicate that absolute AMSA values may be of value beyond the first shocks, whereas the use of AMSA-changes seems restricted to the early phase.
Introduction
Ventricular fibrillation (VF) is the first observed cardiac rhythm in about one third of all out-of-hospital cardiac arrests (OHCAs). Despite all improvements in the chain of care, survival to hospital discharge is still only 5-20%.1-4 Based on the principle that defibrillation on so-called “coarse” VF is more successful than on “fine” VF, analysis of the VF-waveform was introduced. In this context, the amplitude spectrum area (AMSA) has become a topic of increasing interest. 5
Recently, a series of studies on the VF-waveform demonstrated the impact of AMSA in the prediction of outcome. It has repeatedly been demonstrated that the absolute AMSA value predicts not only shock success, but also neurologically favourable outcome.6-8 Moreover, a randomised controlled trial is currently being conducted, investigating shock timing according to the absolute AMSA value, with prolonged chest compressions and delay of compressions if the AMSA considered too low. 9
More recently, changes in AMSA have been implicated to play a role of importance as well. Given the correlation of AMSA with coronary perfusion and myocardial energy levels, an increasing AMSA might reflect an improvement of the myocardial metabolic state.10-12 Recent analyses on human data confirmed that increases of the AMSA are associated with favorable outcome. 7, 13
Currently, most information on the impact of AMSA, and particularly its changes, has been derived from studies investigating the early resuscitation phase.7, 14 The largest studies to date included patients with a median number of two shocks.5, 6 However, the amount of defibrillations is known to vary greatly per patient, and recent studies describe that about half of all patients need more than three shocks in the prehospital setting. 15, 16 At present, little information is available on how AMSA evolves over a longer period of resuscitation, and how these changes relate to shock success.
In view of the above, it seems prudent to improve our understanding of the VF-waveform and its relation with shock success over a longer period of resuscitation. In a cohort including prolonged
resuscitations, we therefore assessed AMSA and AMSA-changes in relation to shock success for both the early and late phase of resuscitation.
Methods
Study population
Prospective registry of OHCA-patients in the Nijmegen area, The Netherlands (≥18 years, non-traumatic arrests), of which details have been described earlier.15 For the current study, we included all patients with VF as first observed cardiac rhythm, who had at least one defibrillation attempt on VF in the prehospital setting. To ensure comparability of the analysed shocks, we excluded patients who received AED shocks prior to arrival of the emergency medical services (EMS). Further exclusion criteria were the absence of either electrocardiogram (ECG) or shock outcome data. Given the observational design, for the present study written informed consent was not necessary to obtain according to the Dutch Act on Medical Research involving Human Subjects.
Emergency medical services
The Nijmegen area (Gelderland-Zuid) has a population of about 530,000 residents and covered 1,040 square kilometres, including urban, suburban and rural areas. The EMS system is activated by calling 112. Paramedics give instructions to the caller to initiate basic life support (BLS), and at least one ambulance is dispatched to the location of the emergency. CPR delivery was performed by EMS personnel according to the prevailing guidelines at the time of arrest. A mechanical chest
compression device (Autopulse®) was part of the standard EMS-equipment, but not routinely used. As it was an observational study , ambulance personnel were not instructed to interrupt chest compressions.
Data acquisition
Demographic and arrest characteristics were collected using EMS and hospital records, according to the Utstein style definitions.17 ECG and transthoracic impedance (TTI) data (sample frequency of 125
Hz and 61 Hz respectively) were recorded with the defibrillator paddles. During the study period, all patients were treated with defibrillators of the same manufacturer (Lifepak, Physio Control,
Redmond, WA, USA), with identical signal processing characteristics (e.g. built-in bandpass filters). The ECG and TTI data were visualised with Codestat (Version 7.0, Physio Control, Redmond, WA, USA) and exported to Matlab-files (.mat) for signal analysis.
VF waveform analysis
VF-waveform analysis was performed with Matlab (Version R2014b, Mathworks, Natick, USA). In this per-shock analysis, we performed analyses at different time points during the resuscitation, i.e. analyses of all first shocks, second shocks, third shocks etc. We specified a minimum of 10 analysable registrations per time point. VF-waveform segments were selected prior to shock delivery, visually free of (chest compression) artefacts and occurring within 30 seconds prior to the shock. Signals were pre-processed with a fourth-order Butterworth bandpass filter with cutoff frequencies of 2 and 48 Hz. A three-second segment of VF (N=376 samples) was used for calculation of the VF-characteristics. All VF-segments were assessed in a uniform manner, blinded for shock outcome.
A discrete fast Fourier transform was performed on all 376 data points for conversion to the
frequency domain. The AMSA was calculated from the obtained frequency spectrum, as the summed product of individual frequencies and their corresponding amplitudes over an interval from 4 to 48 Hz. A more detailed description of the AMSA can be found in Supplement 1.
Furthermore, we calculated the relative difference of the AMSA compared to the previous shock delivery (e.g. the relative change of AMSA between shock 1 and shock 2 is calculated as ∆AMSA = AMSAshock 2/AMSAshock1).
Study groups
Shocks were categorised as either shocks in the early phase, or shocks in the late phase of resuscitation. Dichotomization was based on the median number of shocks in our population.:
Early shocks: Shocks earlier than the median number of shocks in our population
Late shocks: Shocks later than, or equal to, the median number of shocks in our population.
End point and aim of the study
The primary end point of the study was return of organised rhythm (ROOR), defined as at least 2 QRS complexes within 5 seconds, within 60 seconds after shock delivery. We analysed the AMSA and the ∆AMSA in relation to shock success, separately for the early phase (shocks 1 to 3) and the late phase of resuscitation (shocks 4 to 8).
Statistics
Categorical data were reported as numbers (percentages). Continuous variables were analysed for Gaussian distribution and reported as means ± standard deviations or medians (interquartile ranges, IQR), whichever appropriate. Continuous values of AMSA and ΔAMSA were reported as medians with IQR and compared between successful and unsuccessful shocks using the Mann Whitney U test. VF-characteristics were subsequently divided into quartiles and compared in relation to shock success using Chi-square tests for trend. For all tests, we considered a p-value of <0.05 as statistically
significant. All statistical analyses were performed using IBM SPSS statistics software (Version 22, IBM Corp., Armonk, NY, USA).
Results
Study population
Baseline characteristics of the study population are reported in Table 1. We studied a total of 139 OHCA-patients with VF as first observed rhythm. Main reason for exclusion was the absence of defibrillator ECG recordings (Supplementary Figure 1). Mean age of the study population was 63 ± 14 years, 71% (98/139) was male. Of all patients, 76% (97/128) had a witnessed arrest, 60% (81/135) received bystander CPR, 45% (61/137) had sustained ROSC at hospital transportation and 24% (32/132) survived until hospital discharge.
Shock characteristics
The median number of shocks delivered per patient was 4 (IQR 2-7). An overview of in- and excluded shocks can be found in Supplementary Figure 1. In total, 448 shocks were available for analysis, of which 48% was successful (216/448). A total of 286 analysed shocks was delivered in the early phase of resuscitation, of which 46% was successful (131/286), compared to 162 shocks in the late phase, of which 52% (85/162) was successful, p=0.18.
Amplitude spectrum area: early vs. late phase of resuscitation
The median AMSA of all shocks in the early phase was 10.5 mVHz [IQR 6.7-14.5]. The median AMSA was higher prior to successful vs. unsuccessful shocks (11.9 mVHz [7.9-15.9] vs. 8.1 mVHz [5.1-11.1], p<0.001, Table 2). The proportion shock success was 20% in the lowest AMSA quartile and increased across quartiles to 70% in the highest AMSA quartile (p for trend <0.001, Figure 1).
The median AMSA of all shocks in the late phase was 7.7 mVHz [IQR 5.7-11.9]. There was a trend towards a higher AMSA prior to successful vs. unsuccessful shocks (8.6 mVHz [6.0-12.3] vs. 6.8 mVHz
[5.5-11.0], p=0.065, Table 2). The proportion of shock success ranged between 40 and 65% among AMSA quartiles (p for trend = 0.041) (Figure 2).
Figure 3 shows the median values per shock of the AMSA over the course of resuscitation, separately for successful and unsuccessful shocks.
Changes of the AMSA: early vs. late phase of resuscitation
The median ΔAMSA of all shocks in the early phase was 1.06 [IQR 0.89-1.37]. The median ΔAMSA was higher for successful vs. unsuccessful shocks (1.17 [0.94-1.42] vs. 1.02 [0.84-1.25], p=0.018, Table 2). The proportion of shock success ranged between 40 and 65% among ΔAMSA quartiles (p for trend = 0.034, Figure 1).
The median ΔAMSA in the late phase was 0.93 [IQR 0.79-1.10], without significant differences between successful and unsuccessful shocks (p=0.478, Table 2). The proportion of shock success did not differ significantly among ΔAMSA quartiles (p for trend = 0.480, Figure 2).
Discussion
We investigated the VF-waveform in relation to shock success in a real-world cohort of OHCA-patients, thereby separately focusing on the early and late phase of resuscitation. Interestingly, the chance of shock success was comparable in both phases of resuscitation. We demonstrated an association between pre-shock AMSA and shock success in both phases of resuscitation. Notably, this association appeared more strongly in the early phase, when compared to the late phase, where the association between AMSA and shock success was less markedly pronounced. Changes of the AMSA were also positively associated with shock success, but only in the early phase of resuscitation. The current findings implicate that pre-shock AMSA can be used as an indicator of shock success throughout resuscitation, but the use of waveform changes seems restricted to the early phase. Given that a substantial amount of OHCA-patients require more than three shocks in the prehospital setting, there is need for other factors determining shock success in these patients.
Previous studies
Animal studies: In animal studies investigating the VF waveform, a higher AMSA has been related to (defibrillation) outcome, based on its relation with arrest duration.18, 19 The time evolution of VF-waveform characteristics has also been described in experimental studies. In general, after inducing VF in animal hearts, the VF-waveform tends to develop from “coarse” to “fine”; a mechanism that is slowed when CPR is applied to the fibrillating heart.20 Interestingly, in two porcine studies comparing a CPR-first resuscitation strategy with a conventional shock-first strategy, the subsets with CPR showed increasing frequencies of VF and had a higher shock success rate.21, 22 Although we did not analyse the effects of CPR on AMSA in our cohort, our findings were similar in that we found an association between successful shocks and an increasing AMSA in the early phase of cardiac arrest.
Most experimental studies focus on VF in the early phase of cardiac arrest, hampering comparisons with our analyses in the late phase.
Human studies: A recent study investigating the AMSA prior to all shocks in the prehospital setting, demonstrated that the averaged AMSA over the total resuscitation is a strong prognosticator of survival.6 The results of that study already implicated that the AMSA is an important predictor for outcome throughout all phases of resuscitation. Two previous observational studies described individual shock success later than the first defibrillation attempt in VF cardiac arrests. The largest study to date distinguished first and subsequent shocks, and demonstrated that AMSA is predictive for shock success in both shock types.23 It has to be noticed that the abovementioned studies studied a population with a median of two shocks per patient, which is markedly lower compared to our population. Presumably, their results were mainly driven by shocks which we would have categorised as “early”. Therefore, the predictive ability of AMSA for shock success in this cohort remains unknown for what we call the “late phase” of cardiac arrest. A smaller OHCA-study investigating a mean number of 4.4 subsequent shocks per patient, demonstrated a good predictive ability of the AMSA, also in subsequent shocks. 8 As opposed to the latter study, which studied VF-termination, we used ROOR as our definition of shock success, which is a more robust end point. Our findings suggest that there is an association between AMSA and shock success, even in the late phase of resuscitation. Whether the AMSA can consequently be used as a predictor of shock success in late phases of cardiac arrest, should be investigated in future larger studies.
Interestingly, the abovementioned studies distinguished subsequent shocks according to the success of the previous shock. If the patient has shock success, this may influence the probability of shock success of the next shock, and may therefore be a confounding variable. As it was our primary aim to assess the association between AMSA and shock success in early vs. late phase of resuscitation, we did not investigate account for this so-called “recurrent” and “shock-resistant” VF.
Waveform changes
Although studies on the evolving waveform are scarce, VF changes have been increasingly investigated over the past years. A study on OHCA patients in 1993 first described the potential benefits of increasing VF amplitudes for the chances of survival.24 A more recent study on the course of the AMSA showed that increases between the first three defibrillation attempts are independently associated with shock success after OHCA.7 Although the latter study did not investigate the course of VF later than the third shock in a population with very high survival rates, the results are in line with our observations.
Other recent studies on the course of the AMSA demonstrated that taking into account ΔAMSA-measures improves predictive models for return of spontaneous circulation.13, 25 Notably, these studies incorporated resuscitations with a median of 1 to 2 defibrillation attempts per patient. The described AMSA changes of these articles therefore mainly involved the first two shocks, which is especially the phase of resuscitation where we also demonstrated benefits of an increasing AMSA.
In summary, our observations that increases of the AMSA occur between early shock deliveries and indicate a higher likelihood of defibrillation success, are in line with previous studies on this topic, and support the hypothesis that VF-changes are a sign of improving myocardial metabolic state in the early phase of resuscitation. Shocks later than the third shock have scarcely been investigated for waveform changes. Our results indicate that VF-waveform characteristics develop from “coarse” to “fine” VF over the course of resuscitation, but that AMSA-changes are not associated with shock success in the late phase of cardiac arrest. As opposed to the early phase, later phases of resuscitation are typically characterised by a more varying course in terms of refibrillation, the number of shocks delivered and intermittent (perfusing) organised rhythms. Potentially, in this late phase of resuscitation, these factors become more important than changes in the VF-waveform. Moreover, there remains uncertainty about the impact of underlying heart disease on the VF-waveform and its clinical implications with regard to shock outcome. 14, 26-29 Potentially, the presence
or persistence of a myocardial infarction influences the VF-waveform and its relation to shock success in later phases of resuscitation.
Implications
In follow-up on previous studies, we demonstrated that the chance of shock success is comparable in the early and late phase of VF cardiac arrest. Throughout resuscitation, shocks result in organised rhythm in about 50% of the cases. Remarkably, the association of AMSA with shock success is very strong in the early phase of cardiac arrest. In later phases, the AMSA is also associated with shock success, although differences between successful and unsuccessful shocks seemed less pronounced. Appreciating that a substantial amount of VF-patients receive “late” shocks in the prehospital setting, and given that each additional shock worsens prognosis, factors associated with shock success are important in these patients.30 To improve current prediction models, future research should focus on factors other than the VF-waveform that are associated with shock success in this patient group.
Secondly, our findings confirm that increases of the AMSA in the early phase of resuscitation are a good prognostic sign. Therefore, one could implicate that the AMSA is a good marker of myocardial metabolic state and increasing waveform measures represent more “readiness” for defibrillation in the early phase of resuscitation. In the late phase of resuscitation, the VF-waveform tends to develop from “coarse” to “fine”, reflecting a decrease of myocardial metabolic state. In this phase, changes were not associated with shock success, implicating that benefits of AMSA-changes are restricted to the early phase of VF. With regard to trials investigating VF-waveform guided defibrillation strategies, it can be hypothesised that the potential benefits of VF-changes should be made in the early phase of VF. In the late phase, other factors potentially gain more importance.
Limitations
Although we analysed patients from a prospective registry, only patients with analysable VF data and known shock success were included. However, no important differences in baseline characteristics between in- and excluded patients were observed.
Second, in analogy to previous reports, we chose our shock success definition as return of organised rhythm. It has to be noticed that ROOR does not always imply return of spontaneous circulation. Therefore, shocks considered as “successful” might in some cases represent a patient with a pulseless electrical activity.
Conclusion
In the early phase of VF cardiac arrest, both a high AMSA and an increase in AMSA indicated a high likelihood of a successful defibrillation. Although less pronounced, the association between AMSA and shock success pertained in the late phase of resuscitation. Contrastingly, no association between AMSA changes and shock success was found in the late phase of resuscitation. The current findings implicate that AMSA can be used as predictor of shock success throughout the resuscitation, but the use of VF-waveform changes seems restricted to the early phase of resuscitation. Future studies are warranted investigating determinants of shock success in patients with prolonged VF cardiac arrests.
Conflicts of Interest
Prof. de Boer is a member of the European advisory board on interventional cardiology of Medtronic. Prof. van Royen received research grants from Abbott, Biotronik, AstraZeneca and Philips, and
professional fees from Abbott and Medtronic. The other authors have no conflicts of interest to declare.
Tables and Figures
Table 1: Baseline characteristics of the study population. Values are reported as means ± standard deviations,
medians (interquartile ranges), or n (%). CPR = cardiopulmonary resuscitation; EMS = emergency medical services; OHCA = out-of-hospital cardiac arrest; ROSC = return of spontaneous circulation. Baseline characteristics were missing for the following variables: Public location (n=136), witnessed arrest (n=128), bystander CPR (n=135) response time (n=111), number of EMS-shocks (n=138), amiodarone (n=115), adrenaline (n=116), any ROSC (n=136), sustained ROSC (n=137) and survival to discharge (n=132).
OHCA-patients (N=139)
Patient characteristicsAge (years) 63 ± 14
Male gender 98 (71)
Arrest characteristics
Public location arrest 58 (43)
Witnessed arrest 97 (76)
Bystander 93 (73)
EMS 4 (3)
Bystander CPR 81 (60)
Response time (minutes) 8 (5-10) Number of EMS shocks 4 (2-7)
Amiodarone 63 (55)
Adrenaline 95 (82)
Outcome characteristics Any ROSC during resuscitation Sustained ROSC
81 (60) 61 (45)
Table 2: The amplitude spectrum area (AMSA) in relation to shock success (return of organised rhythm). Shocks
were divided in early shocks (shocks 1-3, N=282) and late shocks (shocks 4-8, N=162). Both pre-shock AMSA and the ΔAMSA (change of AMSA compared to previous shock) were reported. Values were reported as medians (interquartile ranges). VF=ventricular fibrillation.
VF-waveform in relation to shock success : Early vs. late phase of resuscitation
Shock success
No shock success
p-value
All shocks
AMSA
9.7 (7.0-14.0)
7.5 (5.3-11.1)
<0.001
Δ AMSA
0.98 (0.86-1.23)
1.00 (0.79-1.17)
0.246
Shocks 1-3
AMSA
11.9 (7.9-15.9)
8.1 (5.1-11.1)
<0.001
Δ AMSA
1.17 (0.94-1.42)
1.02 (0.84-1.25)
0.018
Shocks 4-8
AMSA
8.6 (6.0-12.3)
6.8 (5.5-11.0)
0.065
Δ AMSA
0.92 (0.80-1.07)
0.95 (0.76-1.13)
0.478
Figure 1 – Quartiles of the amplitude spectrum area (AMSA, left) and the change in AMSA (dAMSA, right) in
relation to shock success (return of organised rhythm), for the first three shock deliveries
Figure 2 – Quartiles of the amplitude spectrum area (AMSA, left) and the change in AMSA (dAMSA, right) in
relation to shock success (return of organised rhythm), for shocks later than the third.
Figure 3– Course of the amplitude spectrum area in relation to shock success.
Appendix
Supplement 1 – Pre-processing steps, definitions and mathematical formulation of the amplitude spectrum area (AMSA).
The ventricular fibrillation (VF) signal was sampled at a sampling frequency of fs=125 Hz. Electrocardiographic (ECG) recordings were pre-processed using a fourth-order Butterworth bandpass filter with cut-off frequencies of 2 and 48 Hz. To cancel phase shift the filter was applied once forward and once backward in time, which is what the MATLAB command filtfilt does. The
characteristics were determined from a three-second chest compression free segment of initially-recorded VF ( xn, n=0,1, … , N −1¿ , with the number of samples of the filtered time segment
N=376 . A standard Fast Fourier Transform (FFT) of the VF segment ( xn ) returns N Fourier coefficients ^x0, ^x1, … , ^xN −1 . Only the first N /2 coefficients are relevant and they correspond to the N /2 frequencies fk=k ∙ fs/N with k =0,1, … , N /2 . The amplitude
spectrum area (AMSA) is defined as
2
N
∑
k=0 N −1|
^
x
k∙ f
k|
, but in this sum only those indices k areSupplementary Tables and Figures
Supplementary Figure 1 – Flowchart of the patient population. AED = automated external defibrillator; EMS =
emergency medical services; VF=ventricular fibrillation
VF as first observed rhythm N = 273 patients
AED use 54
Eligible patients N = 219 patients
No analyzable ECG tracing 80
Inclusion N = 139 patients
Shocks
N = 591 EMS shocks ECG without VF segment < 30s prior to shock 107
No shock outcome determinable 17
< 10 registrations 19 Included shocks N = 448 shocks Early phase N=286 shocks Late phase N=162 shocks
Supplementary Table 1: Per-shock analysis of the amplitude spectrum area (AMSA) in relation to shock
success (return of organised rhythm).
Amplitude spectrum area
Shock success
No shock success
p-value
Shock
1
10.3 (4.9-14.1)
7.5 (5.0-10.5)
0.0012
12.6 (9.1-17.0)
8.6 (4.8-11.3)
<0.0013
12.0 (7.8-15.9)
9.1 (5.3-12.8)
0.034
10.6 (7.5-13.1)
7.6 (5.5-13.6)
0.175
9.6 (8.2-12.4)
6.7 (5.3-8.9)
0.0066
7.6 (5.9-10.0)
6.1 (4.9-12.4)
0.4177
6.0 (4.9-9.3)
7.0 (5.6-12.3)
0.3088
6.9 (5.3-12.0)
6.6 (5.6-8.7)
0.591References
[1] Berdowski J, Berg RA, Tijssen JG, Koster RW. Global incidences of out-of-hospital cardiac arrest and survival rates: Systematic review of 67 prospective studies. Resuscitation 2010;81:1479-87.
[2] Weisfeldt ML, Sitlani CM, Ornato JP, et al. Survival after application of automatic external defibrillators before arrival of the emergency medical system: evaluation in the resuscitation outcomes consortium population of 21 million. J Am Coll Cardiol 2010;55:1713-20.
[3] Daya MR, Schmicker RH, Zive DM, et al. Out-of-hospital cardiac arrest survival improving over time: Results from the Resuscitation Outcomes Consortium (ROC). Resuscitation 2015;91:108-15. [4] Nas J, Thannhauser J, Herrmann J, et al. Changes in Automated External Defibrillator Use and Survival after Out-of-Hospital Cardiac Arrest in the Nijmegen Area. Netherlands Heart Journal 2018. [5] Ristagno G, Mauri T, Cesana G, et al. Amplitude spectrum area to guide defibrillation: a validation on 1617 patients with ventricular fibrillation. Circulation 2015;131:478-87.
[6] Indik JH, Conover Z, McGovern M, et al. Association of amplitude spectral area of the ventricular fibrillation waveform with survival of out-of-hospital ventricular fibrillation cardiac arrest. J Am Coll Cardiol 2014;64:1362-9.
[7] Schoene P, Coult J, Murphy L, et al. Course of quantitative ventricular fibrillation waveform measure and outcome following out-of-hospital cardiac arrest. Heart Rhythm 2014;11:230-6.
[8] Shanmugasundaram M, Valles A, Kellum MJ, Ewy GA, Indik JH. Analysis of amplitude spectral area and slope to predict defibrillation in out of hospital cardiac arrest due to ventricular fibrillation (VF) according to VF type: recurrent versus shock-resistant. Resuscitation 2012;83:1242-7.
[9] Real Time Amplitude Spectrum Area to Guide Defibrillation. https://ClinicalTrials.gov/show/NCT03237910.
[10] Gazmuri RJ, Kaufman CL, Baetiong A, Radhakrishnan J. Ventricular Fibrillation Waveform Changes during Controlled Coronary Perfusion Using Extracorporeal Circulation in a Swine Model. PLoS One 2016;11:e0161166.
[11] Salcido DD, Menegazzi JJ, Suffoletto BP, Logue ES, Sherman LD. Association of intramyocardial high energy phosphate concentrations with quantitative measures of the ventricular fibrillation electrocardiogram waveform. Resuscitation 2009;80:946-50.
[12] Eftestol T, Wik L, Sunde K, Steen PA. Effects of cardiopulmonary resuscitation on predictors of ventricular fibrillation defibrillation success during out-of-hospital cardiac arrest. Circulation 2004;110:10-5.
[13] Nakagawa Y, Amino M, Inokuchi S, Hayashi S, Wakabayashi T, Noda T. Novel CPR system that predicts return of spontaneous circulation from amplitude spectral area before electric shock in ventricular fibrillation. Resuscitation 2017;113:8-12.
[14] Hulleman M, Salcido DD, Menegazzi JJ, et al. Predictive value of amplitude spectrum area of ventricular fibrillation waveform in patients with acute or previous myocardial infarction in out-of-hospital cardiac arrest. Resuscitation 2017;120:125-31.
[15] Verhaert DV, Bonnes JL, Nas J, et al. Termination of resuscitation in the prehospital setting: A comparison of decisions in clinical practice vs. recommendations of a termination rule. Resuscitation 2016;100:60-5.
[16] van Alem AP, Post J, Koster RW. VF recurrence: characteristics and patient outcome in out-of-hospital cardiac arrest. Resuscitation 2003;59:181-8.
[17] Jacobs I, Nadkarni V, Bahr J, et al. Cardiac arrest and cardiopulmonary resuscitation outcome reports: update and simplification of the Utstein templates for resuscitation registries: a statement for healthcare professionals from a task force of the International Liaison Committee on Resuscitation (American Heart Association, European Resuscitation Council, Australian Resuscitation Council, New
Zealand Resuscitation Council, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Councils of Southern Africa). Circulation 2004;110:3385-97.
[18] Indik JH, Shanmugasundaram M, Allen D, et al. 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. Resuscitation 2009;80:1420-3.
[19] Indik JH, Allen D, Shanmugasundaram M, et al. Predictors of resuscitation in a swine model of ischemic and nonischemic ventricular fibrillation cardiac arrest: superiority of amplitude spectral area and slope to predict a return of spontaneous circulation when resuscitation efforts are prolonged. Crit Care Med 2010;38:2352-7.
[20] Lee JC, Suh GJ, Kim HC. Electrocardiogram frequency change by extracorporeal blood perfusion in a swine ventricular fibrillation model. Biomed Eng Online 2013;12:123.
[21] 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;40:563-70.
[22] 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;32:1352-7.
[23] Ristagno G, Li Y, Fumagalli F, Finzi A, Quan W. Amplitude spectrum area to guide resuscitation-a retrospective analysis during out-of-hospital cardiopulmonary resuscitation in 609 patients with ventricular fibrillation cardiac arrest. Resuscitation 2013;84:1697-703.
[24] Callaham M, Braun O, Valentine W, Clark DM, Zegans C. Prehospital cardiac arrest treated by urban first-responders: profile of patient response and prediction of outcome by ventricular fibrillation waveform. Ann Emerg Med 1993;22:1664-77.
[25] He M, Lu Y, Zhang L, Zhang H, Gong Y, Li Y. Combining Amplitude Spectrum Area with Previous Shock Information Using Neural Networks Improves Prediction Performance of Defibrillation Outcome for Subsequent Shocks in Out-Of-Hospital Cardiac Arrest Patients. PLoS One 2016;11:e0149115.
[26] Hidano D, Coult J, Blackwood J, et al. Ventricular fibrillation waveform measures and the etiology of cardiac arrest. Resuscitation 2016;109:71-5.
[27] Olasveengen TM, Eftestol T, Gundersen K, Wik L, Sunde K. Acute ischemic heart disease alters ventricular fibrillation waveform characteristics in out-of hospital cardiac arrest. Resuscitation 2009;80:412-7.
[28] Bonnes JL, Thannhauser J, Hermans MC, et al. Ventricular fibrillation waveform characteristics differ according to the presence of a previous myocardial infarction: A surface ECG study in ICD-patients. Resuscitation 2015;96:239-45.
[29] Bonnes JL, Keuper W, Westra SW, et al. Characteristics of ventricular fibrillation in relation to cardiac aetiology and shock success: A waveform analysis study in ICD-patients. Resuscitation 2015;86:95-9.
[30] Hasegawa M, Abe T, Nagata T, Onozuka D, Hagihara A. The number of prehospital defibrillation shocks and 1-month survival in patients with out-of-hospital cardiac arrest. Scandinavian journal of trauma, resuscitation and emergency medicine 2015;23:34.
