- Email: [email protected]
Acute ischemic heart disease alters ventricular fibrillation waveform characteristics in out-of hospital cardiac arrest
Resuscitation 80 (2009) 412–417
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical Paper
Acute ischemic heart disease alters ventricular fibrillation waveform characteristics in out-of hospital cardiac arrest夽 Theresa M. Olasveengen a,∗ , Trygve Eftestøl b , Kenneth Gundersen b , Lars Wik c , Kjetil Sunde a a
Institute for Experimental Medical Research and Department of Anaesthesiology, Ulleval University Hospital, N-0407 Oslo, Norway Department of Electrical and Computer Engineering, Faculty of Science and Technology, University of Stavanger, Stavanger, Norway c The National Competence Centre for Emergency Medicine and Institute for Experimental Medical Research, Ulleval University Hospital, N-0407 Oslo, Norway b
a r t i c l e
i n f o
Article history: Received 20 August 2008 Received in revised form 12 January 2009 Accepted 19 January 2009 Keywords: Advanced life support Cardiac arrest Electrocardiography Heart failure Ischemia Myocardial infarction Out-of-hospital CPR Ventricular fibrillation
a b s t r a c t Background: Although ventricular fibrillation waveform characteristics (VFWC) correlate with coronary perfusion pressure and may predict defibrillation outcome, recent animal data indicate that these waveform characteristics are altered in both acute myocardial infarction (AMI) and chronic coronary heart disease (CHD). We wanted to confirm these recent animal data in humans and explore the possibility for such characteristics to identify acute ischemia during cardiac arrest. Methods: Data from all adult patients admitted to hospital after out-of-hospital VF cardiac arrest in Oslo between May 2003 and July 2007 were prospectively collected. Patients were categorized into one of four pre-defined etiologic groups: patients with AMI (AMI only), patients with AMI and CHD (AMI and CHD), patients with previous CHD without evidence for a new AMI (CHD only), and patients with primary arrhythmia (PA). VFWC were analyzed from prehospital ECG tracings, and the different etiologic groups compared using ANOVA. Results: One-hundred-and-one patients with ECG recordings usable for VF analysis could confidently be categorized; 16 with AMI only, 34 with AMI and CHD, 41 with CHD only and 10 with PA. The two VFWC median slope (MS) and amplitude spectral area (AMSA) were significantly depressed in patients with AMI only compared to both PA (MS p = 0.008, AMSA p = 0.035) and CHD only patients (MS p = 0.008, AMSA p = 0.006). Conclusions: AMI patients have depressed MS and AMSA compared to patients without AMI during VF cardiac arrest. VFWC might be helpful in identifying patients with AMI during cardiac arrest, but prospective clinical studies are warranted to assess its feasibility and clinical benefit. © 2009 Elsevier Ireland Ltd. All rights reserved.
Introduction Approximately 30–40% of all out-of-hospital cardiac arrest (OHCA) patients are found with initial ventricular fibrillation (VF),1–3 and typically ∼20% of these patients survive to hospital discharge.3,4 It can be expected that 60–80% of the OHCA patients with cardiac aetiology have underlying acute or chronic ischemic heart disease.5,6 When angiography was performed on consecutive patients admitted to our hospital after OHCA with presumed cardiac aetiology ∼50% had acute myocardial infarction (AMI).7 There is wide consensus that early revascularization with percutaneous coronary intervention (PCI) or fibrinolytic therapy will
夽 A Spanish translated version of the summary of this article appears as Appendix in the final online version at doi:10.1016/j.resuscitation.2009.01.012. ∗ Corresponding author at: Institute for Experimental Medical Research, Ullevål University Hospital, University of Oslo, Division Ullevål University, Hospital, N-0407 Oslo, Norway. Tel.: +47 23016837/41419930; fax: +47 23016799. E-mail address: [email protected] (T.M. Olasveengen). 0300-9572/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.resuscitation.2009.01.012
decrease mortality for patients with ST-segment elevation myocardial infarction (STEMI),8,9 and that time from ischemic event to reperfusion is crucial.10,11 The same strategy has been advocated for cardiac arrest patients after return of spontaneous circulation (ROSC) when AMI is suspected.7,12–14 However, only approximately half of cardiac arrest patients with initial VF and presumed cardiac aetiology can be expected to achieve sustained ROSC during advanced life support (ALS),15 and prolonged resuscitation efforts are often ultimately terminated in the field with little else to offer these patients. Is has recently been demonstrated that coronary angiography and PCI is possible during continuous mechanical chest compressions,16,17 and may be a corrective treatment for a subgroup of cardiac arrest patients where ROSC is not achieved with traditional ALS. Patients with AMI and shock resistant VF cardiac arrest, but with “hearts and brains too good to die”, could be transported with mechanical CPR directly to angiography where underlying occlusions could be treated. Such a strategy could shorten the all important time to reperfusion and perhaps save lives of patients that would otherwise be terminated in the field. A major
T.M. Olasveengen et al. / Resuscitation 80 (2009) 412–417
challenge is early identification of patients who would benefit from this aggressive treatment. VF waveform analysis correlates with coronary perfusion pressure18 and has been described as a method to predict when a VF is likely to be shocked into a pulse generating rhythm, thereby reducing pauses in chest compressions and potential myocardial damage by limiting redundant shocks.19,20 It was recently shown that animal VF waveform characteristics appear altered after acute myocardial infarction (AMI)21 and chronic heart failure,22 and it may therefore be questionable to extrapolate earlier animal data where cardiac arrest has been induced electrically on structurally normal hearts. We are unaware of any studies on the effects of ischemic heart disease on human VF waveform characteristics. We hypothesize that VF waveform characteristics in patients with cardiac arrest and AMI and/or chronic coronary heart disease (CHD) will be different compared to patients with arrhythmic VF-arrest, and that VF analysis may contribute to identify ischemic conditions with indications for reperfusion therapy during cardiac arrest.
413
impedance changes. Annotations were made while reviewing available clinical information from the Utstein forms and ambulance records. Total time without spontaneous circulation (CPR time), time without chest compressions divided by CPR time (Hands-off ratio), compression rate and the actual number of compressions and ventilations per minute were calculated for each episode. Patient classification Patients were classified into the following four etiologic categories: (1) AMI without any previous symptoms or treatment for CHD (AMI only), (2) AMI in patients with previous symptoms or treatment for CHD (AMI and CHD), (3) Patients with previous symptoms or treatment for CHD without AMI (CHD only), (4) Primary arrhythmia without any previous symptoms or treatment for CHD (PA),
Material and methods Study design and recruitment All patients older than 18 years admitted to hospital after non-traumatic VF OHCA between May 2003 and July 2007 were retrospectively studied. Due to an ongoing randomized study of the effect of intravenous access and drugs in the Oslo Emergency Medical Service (EMS), some of our included patients are also included in this study registered at clinicaltri.gov (NCT00121524). Approximately half of the patients will therefore be expected to have received intravenous drugs sometime during the resuscitation effort. As we have used the first seconds of the ECG registration in our analysis (see VF characteristics below), randomization to the following protocols including or excluding intravenous drugs should be of little importance. The EMS system in Oslo is a one-tiered centralized community run system for a population of 540 000. In weekdays between 7:30 AM and 10 PM a physician-manned ambulance staffed by two paramedics and an anaesthesiologist functions on the same level as the regular paramedic staffed ambulances.
In addition, we registered whether the patients had been diagnosed with chronic heart failure prior to the arrest (typical clinical signs, use of heart failure medication and ejection fraction <50%, if available). The following information was considered when evaluating aetiology and degree of ischemia (from highest to lowest importance): autopsy reports, coronary intervention reports, previous medical history, biochemistry (including troponin release), decision to install an Implantable Cardioverter Defibrillator (ICD), and ECG findings. The AMI diagnosis was only used if there were clear angiographic or autopsy descriptions, or if there was a substantial and increasing Troponin release in the following first three blood samples. Small troponin release alone (<0.5 mcg/l) was not classified as AMI. Cases where the patient record did not contain sufficient information to determine the causes of cardiac arrest were excluded. The classifications were done independently by two physicians with extensive experience in prehospital ECG annotations/rhythm interpretation (TMO and KS), and any eventual differences in opinion were discussed until consensus was reached. Importantly, this classification was done independently and at a different location than the VF waveform calculations (TE and KG).
Data collection VF characteristics Utstein forms are routinely filled out by ambulance personnel after every cardiac arrest and submitted to the study supervisor along with a copy of the ambulance run sheet. Automated, computer based time records from the dispatch centre supplement ambulance run sheets with regard to response times. For all admitted patients, additional hospital records are obtained from the respective receiving hospitals. Information from Utstein forms, ambulance run sheets, dispatch and hospital records are linked together with continuous ECG tracings as described below. Equipment and data processing Standard LIFEPAK 12 defibrillators (Physio-Control, a Division of Medtronic, Redmond, WA, USA) were used, which routinely measure transthoracic impedance by applying a near constant sinusoidal current across the standard defibrillation pads. After a CPR effort the ECGs with transthoracic impedance signals were normally transferred to a local server at The National Competence Centre for Emergency Medicine (Ulleval University Hospital, Oslo, Norway), and data from each case were viewed and annotated using a CODE-STATTM 7.0 (Physio-Control, Redmond, WA, USA) for detection of chest compressions and ventilations from transthoracic
The data files for the included patients were acquired from the CODE-STAT system using a proprietary software program provided by Physio-Control for extracting the data. The extracted data files include defibrillator log data of important events, sampled measurements of electrocardiograms (ECG) and impedance and manual rhythm annotations. The sample rates of the ECG and impedance channels are 125 Hz and 61 Hz respectively. The log data, ECG, impedance data and rhythm annotations were integrated in MATLAB (The Mathworks Inc.) where further signal analysis was performed. All initial VF periods without compressions were identified. These chest compression pauses were generally pauses for on-scene initial rhythm assessment. The ECG tracings from these periods were extracted, and six different VF characteristics were computed for the last 125 samples of all identified tracings (corresponding to 1 s duration of the analysis window). Five previously studied VF characteristics based on analysis of spectral characterization were chosen. Prior to calculating the VF characteristics themselves the spectral representation of the signal segment to be analyzed is represented by its fast Fourier transform giving the amplitude and corresponding frequency of each of the signal’s frequency compo-
414
T.M. Olasveengen et al. / Resuscitation 80 (2009) 412–417
nents. The VF characteristics are derived from the signal’s spectral representation and were as follows: 1. The amplitude spectrum relationship (AMSA) is computed as the sum of weighted amplitude values weighted by frequency value, thus emphasizing high frequency content.23,24 2. Centroid frequency (CF) characteristics is computed as the frequency corresponding to the point of mass in the spectrum.25 Peak power frequency (PPF) corresponds to the frequency of the maximum spectral amplitude.25 The spectral flatness measure (SFM) is computed as the ratio of the geometric mean and the arithmetic mean of the spectrum.20 3. The signal energy (ENRG) is computed as the area under the spectrum.20 4. The median slope (MS) is computed as the median value of amplitude differences in the analysis window, thus also emphasizing the high frequency content.26 AMSA and MS combine information about amplitude and frequency so that high frequency and amplitude VF (coarse VF) give high AMSA and MS values. Statistical analysis Statistical calculations were performed using a spreadsheet program (Excel 2002, Microsoft Corp, Redmond, WA, USA) or a statistical software package (SPSS 14.0, SPSS Inc., Chicago, IL, USA). Natural logarithm was computed for MS, AMSA, CF and ENRG values before statistical analysis to change the distribution of the parameters to be approximately Gaussian. Comparisons between the different etiologic groups were done with ANOVA and independent samples t-test with Bonferroni corrections. p-values less than 0.05 were considered significant. Results From May 2003 to July 2007 the Oslo EMS responded to 1867 cardiac arrests. ALS was attempted in 931 cases, whereof 312 (34%) were found in VF. Among the VF patients, 205 (66%) were admitted to hospital. One-hundred and seven (52%) of the 205 cases eligible for inclusion had usable ECG recordings with impedance signals. Six additional cases had to be excluded as they could not confidently be classified into one of the four defined etiologic categories from available patient information, leaving 101 cases (49%) included for further analysis. General Utstein data with CPR quality parameters are presented in Table 1. There were no significant differences in Utstein characteristics between included patients and the patient group as a whole (Table 1). Among the 101 included patients, 37 had previously been diagnosed and treated for chronic heart failure while 54 had not exhibited clinical signs of chronic heart failure prior to their cardiac arrest (10 unknown). Fifty percent of the patients in the AMI + CHD group and 61% in the CHD group had confirmed chronic heart failure prior to arrest (none in the two other groups). With ANOVA two of the six VF waveforms characteristics varied significantly between the etiologic groups; MS (p = 0.003) and AMSA (p = 0.005) (Table 2). With post hoc Bonferroni corrected independent t-tests both MS and AMSA were significantly depressed for the AMI only group compared to both PA only (p = 0.008 and 0.035, respectively) and CHD only groups (p = 0.008 and 0.006, respectively). The overall trend for the VF waveform characteristics were values increasing from the AMI only → AMI and CHD → CHD only → PA (Table 2). Compared to all patients without AMI, all patients with AMI had significantly depressed MS (4.3 vs. 5.5, p = 0.002), AMSA (7.8
Table 1 Utstein data with CPR quality parameters for out-of-hospital cardiac arrest patients who present with initial ventricular fibrillation and are subsequently admitted to hospital. All VF cardiac arrests admitted to hospital (n = 205)
Included VF cardiac arrests admitted to hospital (n = 101)
Female Age (years) Cardiac aetiology
52 (25) 64 ± 16 186 (91)
22 (22) 66 ± 16 96 (95)
Location of arrest Home Work Public place Other
73 (36) 12 (6) 102 (50) 18 (9)
31 (31) 4 (4) 58 (57) 7 (7)
Cardiac arrest witnessed by Bystanders AED-trained bystanders Ambulance personnel Not witnessed
151 (74) 7 (3) 29 (14) 18 (9)
84 (83) 2 (2) 3 (3) 12 (12)
Bystander CPR Endotracheal intubation Response time (min) Time from arrest to ALS (min) CPR quality Hands-off ratio Compression rate Compressions (min−1 ) Ventilations (min−1 )
132 (64) 152 (74) 7 (5.8) 8 (7.9)
70 (69) 86 (85) 7 (6.8) 8 (7.10)
N/A N/A N/A N/A
0.24 ± 0.14 118 ± 11 88 ± 17 12 ± 4
Transported during ongoing CPR Admitted to ICU Discharged alive
47 (23)
20 (20)
160 (78) 95 (46)
90 (89) 50 (50)
CPC score 1 2 3 4
a
b
79 8 5 0
40 5 4 0
Values given as number of cases with percentage except from age and CPR quality parameters given as mean with standard deviation and response times given as median with 95% confidence intervals. a CPC classification not available for three patients. b CPC classification not available for one patient.
vs. 13.3, p = 0.003) and ENRG (893 vs. 1210, p = 0.025). Initial MS values in patients with and without AMI are shown in Figure 1. The values for the remaining VF waveform characteristics, CF, PPF and FLAT, were also lower in the patients with AMI compared to
Figure 1. Box-plot of ventricular fibrillation (VF) initial median slope (MS) for all patients with acute myocardial infarction (AMI) and all patients without AMI. pValue obtained from an independent t-test.
T.M. Olasveengen et al. / Resuscitation 80 (2009) 412–417
415
Table 2 Description of VF waveform characteristics for the different etiologic groups. VF waveform characteristic
AMI only (n = 16)
AMI + CHD (n = 34)
CHD only (n = 41)
PA (n = 10)
ANOVA (p-value)
Median slope (MS) Amplitude spectral area (AMSA) Energy (ENRG) Centroid frequency (CF) Peak power frequency (PPF) Spectral flatness (FLAT)
3.7 (2.4, 5.5) 6.2 (2.7, 11.7) 781 (339, 1624) 5.1 (4.1, 6.4) 4.3 (2.5, 5.4) 0.086 (0.065, 0.152)
4.6 (3.7, 5.5) 9.4 (4.6, 15.3) 1025 (662, 1466) 5.1 (4.8, 6.3) 4.4 (3.7, 5.9) 0.112 (0.101, 0.126)
4.9 (4.3, 6.1) 11.7 (7.1, 19.3) 1206 (928, 1591) 6.0 (5.0, 6.8) 4.9 (3.7, 5.9) 0.106 (0.078, 0.136)
6.1 (3.7, 9.8) 15.7 (4.5, 50.2) 1840 (551, 2984) 6.7 (3.6, 7.9) 5.8 (3.7, 7.4) 0.144 (0.055, 0.192)
0.003 0.005 0.083 0.505 0.526 0.879
VF waveform characteristics are given as median values (95% confidence intervals). The etiologic groups were compared using ANOVA. Natural logarithm was computed for MS, AMSA, ENRG and CF values before statistical analysis to improve normal distribution.
Table 3 VF waveform characteristics in patients with and without acute myocardial infarction. VF waveform characteristic
Patients with AMI (n = 50)
Patients without AMI (n = 51)
p-Value
Median slope (MS) Amplitude spectral area (AMSA) Energy (ENRG) Centroid frequency (CF) Peak power frequency (PPF) Spectral flatness (FLAT)
4.3 (3.1, 5.5) 7.8 (4.4, 12.0) 893 (703, 1100) 5.1 (4.6, 6.3) 4.4 (4.2, 5.1) 0.107 (0.080, 0.126)
5.5 (4.3, 6.7) 13.3 (7.1, 19.3) 1210 (833, 1798) 6.1 (5.0, 6.9) 4.9 (3.7, 5.9) 0.111 (0.076, 0.152)
0.002 0.003 0.025 0.284 0.520 0.616
Values in patients with and without acute myocardial infarction (AMI) were compared using independent t-tests. Natural logarithm was computed for MS, AMSA, ENRG and CF values before statistical analysis in order to approach a normal distribution.
the patients without AMI—but the differences were not significant (Table 3). Similarly, patients with chronic heart failure tended to have slightly depressed VF waveforms characteristics compared to those without chronic heart failure, although there were no statistically significant differences: AMSA 9.6 vs. 11.7 (p = 0.460), ENRG 1083 vs. 1100 (p = 0.305), CF 5.6 vs. 5.8 (p = 0.850), PPF 4.9 vs. 4.7 (p = 0.677), FLAT 0.12 vs. 0.10 (p = 0.549), respectively. Median MS was 4.9 in both groups. Discussion The present human data confirm recent animal data21,22 showing that ischemic heart disease alters VF waveform characteristics. Acute ischemic events seem to have a larger impact on VF characteristics than chronic ischemic disease and heart failure. VF waveform analysis could therefore potentially be useful in identifying patients with underlying AMI not achieving ROSC with traditional ALS, and this additional information could influence the ALS providers with regard to further treatment strategy. Early revascularization could be an option in patients with refractory VF and low VF waveform characteristics, selecting these patients for either on-scene thrombolytic therapy,27 or transport to the nearest angiographic laboratory for acute PCI utilizing a mechanical chest compression devise.16 Although thrombolytic treatment was not found to increase survival in a large randomized clinical trial of all patients with cardiac arrest with presumed cardiac cause (TROICA),27,28 the subgroup with suspected AMI could still benefit from such treatment.29 Early PCI during mechanical CPR is another strategy that might potentially increase survival for AMI patients with shock resistant VF.16 In order for PCI to be feasible before ROSC, the time from arrest to PCI facility arrival must be relatively short with a mechanical chest compression devise being used to secure adequate perfusion.30,31 PCI during CPR is costly and resource consuming, and such a treatment protocol would be dependent on good selection criteria to identify patients expected to benefit from such treatment.17,32 We tested six different VF waveform characteristics known to have varying shock prediction abilities. Although we find significantly depressed VF waveform characteristics (MS, AMSA and
ENRG) in patients with AMI compared to patients without AMI, the box plot in Figure 1 clearly illustrates that even the most promising characteristic (MS) does not provide a well defined cut-off value for clinical use. These characteristics would in their current form not give conclusive information on the presence of an AMI, but could be combined with other factors such as previous history, symptoms prior to arrest, age, anoxia time to aid ALS providers in their decision-making. Exploring other waveform characteristics, or modifying the characteristics already described, might yield a more clinically useful “AMI indicator” for cardiac arrest patients with shock resistant VF. Increasing research efforts to improve our knowledge and help identify the underlying mechanisms of cardiac arrest has been advocated, and is hoped to provide specific, corrective therapy, treating the underlying cause of arrest.33 The best shock predictors are generally found to be AMSA and MS with sensitivities ∼95% and specificities ∼60% when performed retrospectively.23,24,34 The usefulness of these VF characteristics are still being debated, as they have never been proven useful in a prospective clinical setting. The usefulness of VF analysis during cardiac arrest, both in shock prediction in general and in other applications such as AMI detection, needs to be investigated prospectively. The most important limitation to this study is the patient selection, as only patients admitted to hospital could be classified into our four etiologic groups. Ideally, all patients with initial VF should have been included and classified according to PCI reports, biochemical markers, ECG findings or autopsy results. As only a small fraction of cardiac arrest deaths are autopsied in Oslo, and virtually no patients dying before hospital admission, such information is not available. Exploring relationships between VF characteristics and structural abnormalities in the heart should be explored further, and in order to test the usefulness of VF analysis in OHCA, a prospective randomized trial is warranted. Another limitation for our data is that they are primarily gathered for another ongoing randomized interventional study on the effects of IV drugs during resuscitation which entails that two different treatment strategies are given to these patients. Any potential effect on VF characteristics from IV drugs such as epinephrine or amiodarone could be a confounding factor, but there is no rea-
416
T.M. Olasveengen et al. / Resuscitation 80 (2009) 412–417
son to believe the randomization to be unevenly distributed among the etiologic groups. Although this provides us with prospectively collected data, the present study design is retrospective. Conclusion Patients with ventricular fibrillation after acute myocardial infarction (AMI) have altered waveform characteristics compared to patients without AMI. Further exploration of such characteristics could yield a tool that might be helpful for identifying patients with AMI with the potential of targeted revascularization therapy during ALS. Prospective clinical studies are warranted to assess its feasibility and clinical benefit. Conflict of interest Olasveengen has received honoraria from Medtronic (Oslo, Norway) and research support from Laerdal Medical Corporation (Stavanger, Norway). Eftestol is the project leader and Gundersen a participant of a project granted by the Research Council of Norway where Laerdal Medical amongst others has contributed with data. Wik is on a Medical Advisory Board for Physio-Control, has in the past consulted for Laerdal and Jolife, and is the principle investigator for a multi-centre mechanical chest compression device study sponsored by Zoll. Sunde has no conflicts of interest to declare. Acknowledgements We thank all physicians and paramedics working in the Oslo EMS Service, as well as Martin Samdal for invaluable help in obtaining patient records, and especially Petter Andreas Steen for discussions and critique. The study was supported by grants from Eastern Norway Regional Health Authority, Ulleval University Hospital, Norwegian Air Ambulance Foundation, Laerdal Foundation for Acute Medicine, and Anders Jahres Fund. Software necessary for analysing data from the LIFEPAK 12 defibrillators was provided by PhysioControl (Redmond, WA, USA). Financial support: The study was supported by grants from Eastern Norway Regional Health Authority, Ulleval University Hospital, Laerdal Foundation for Acute Medicine, Anders Jahres Fund, The Research Council of Norway. The necessary software for analyzing data from the LIFEPAK 12 defibrillators was provided by Physio-Control, Redmond, WA, USA. Author contributions: Olasveengen and Eftestøl have full access to generated data and take full responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Olasveengen, Eftestøl, and Sunde. Acquisition of data: Olasveengen and Wik. Analysis and interpretation of data: Olasveengen, Eftestøl, and Gundersen. Drafting of the manuscript: Olasveengen, Eftestøl, and Sunde. Critical revision of the manuscript for important intellectual content: Olasveengen, Eftestøl, Gundersen, Wik, and Sunde. Statistical expertise: Olasveengen, Eftestøl, and Gundersen. Obtained funding: Sunde, Wik, and Eftestøl. Study supervision: Sunde. References 1. Cobb LA, Fahrenbruch CE, Olsufka M, Copass MK. Changing incidence of out-ofhospital ventricular fibrillation, 1980–2000. JAMA 2002;288:3008–13. 2. Herlitz J, Engdahl J, Svensson L, Young M, Angquist KA, Holmberg S. Decrease in the occurrence of ventricular fibrillation as the initially observed arrhythmia after out-of-hospital cardiac arrest during 11 years in Sweden. Resuscitation 2004;60:283–90. 3. Atwood C, Eisenberg MS, Herlitz J, Rea TD. Incidence of EMS-treated out-ofhospital cardiac arrest in Europe. Resuscitation 2005;67:75–80. 4. Rea TD, Eisenberg MS, Sinibaldi G, White RD. Incidence of EMS-treated out-ofhospital cardiac arrest in the United States. Resuscitation 2004;63:17–24.
5. Silfvast T. Cause of death in unsuccessful prehospital resuscitation. J Int Med 1991;229:331–5. 6. Sunde K, Pytte M, Jacobsen D, et al. Implementation of a standardised treatment protocol for post resuscitation care after out-of-hospital cardiac arrest. Resuscitation 2007;73:29–39. 7. Spaulding CM, Joly LM, Rosenberg A, et al. Immediate coronary angiography in survivors of out-of-hospital cardiac arrest. N Engl J Med 1997;336:1629–33. 8. Fibrinolytic Therapy Trialists’ (FTT) Collaborative Group. Indications for fibrinolytic therapy in suspected acute myocardial infarction: collaborative overview of early mortality and major morbidity results from all randomised trials of more than 1000 patients. Lancet 1994;343:311–22. 9. Keeley EC, Boura JA, Grines CL. Primary angioplasty versus intravenous thrombolytic therapy for acute myocardial infarction: a quantitative review of 23 randomised trials. Lancet 2003;361:13–20. 10. Antman EM, Anbe DT, Armstrong PW, et al. ACC/AHA guidelines for the management of patients with ST-elevation myocardial infarction: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Revise the 1999 Guidelines for the Management of Patients with Acute Myocardial Infarction). Circulation 2004;110:E82–292. 11. Van de WF, Ardissino D, Betriu A, et al. Management of acute myocardial infarction in patients presenting with ST-segment elevation. The Task Force on the Management of Acute Myocardial Infarction of the European Society of Cardiology. Eur Heart J 2003;24:28–66. 12. Bendz B, Eritsland J, Nakstad AR, et al. Long-term prognosis after out-of-hospital cardiac arrest and primary percutaneous coronary intervention. Resuscitation 2004;63:49–53. 13. Keelan PC, Bunch TJ, White RD, Packer DL, Holmes Jr DR. Early direct coronary angioplasty in survivors of out-of-hospital cardiac arrest. Am J Cardiol 2003;91, 1461-3, A. 14. Quintero-Moran B, Moreno R, Villarreal S, et al. Percutaneous coronary intervention for cardiac arrest secondary to ST-elevation acute myocardial infarction. Influence of Immediate Paramedical/Medical Assistance on Clinical Outcome. J Invasive Cardiol 2006;18:269–72. 15. Sunde K, Eftestol T, Askenberg C, Steen PA. Quality assessment of defribrillation and advanced life support using data from the medical control module of the defibrillator. Resuscitation 1999;41:237–47. 16. Grogaard HK, Wik L, Eriksen M, Brekke M, Sunde K. Continuous mechanical chest compressions during cardiac arrest to facilitate restoration of coronary circulation with percutaneous coronary intervention. J Am Coll Cardiol 2007;50:1093–4. 17. Larsen AI, Hjornevik AS, Ellingsen CL, Nilsen DW. Cardiac arrest with continuous mechanical chest compression during percutaneous coronary intervention: a report on the use of the LUCAS device. Resuscitation 2007;75:454–9. 18. Noc M, Weil MH, Tang W, Sun S, Pernat A, Bisera J. Electrocardiographic prediction of the success of cardiac resuscitation. Crit Care Med 1999;27:708–14. 19. Xie J, Weil MH, Sun S, et al. High-energy defibrillation increases the severity of postresuscitation myocardial dysfunction. Circulation 1997;96:683–8. 20. Eftestol T, Sunde K, Ole AS, Husoy JH, Steen PA. Predicting outcome of defibrillation by spectral characterization and nonparametric classification of ventricular fibrillation in patients with out-of-hospital cardiac arrest. Circulation 2000;102:1523–9. 21. Indik JH, Donnerstein RL, Berg RA, Hilwig RW, Berg MD, Kern KB. Ventricular fibrillation frequency characteristics are altered in acute myocardial infarction. Crit Care Med 2007;35:1133–8. 22. Indik JH, Donnerstein RL, Kern KB, Goldman S, Gaballa MA, Berg RA. Ventricular fibrillation waveform characteristics are different in ischemic heart failure compared with structurally normal hearts. Resuscitation 2006;69:471–7. 23. Marn-Pernat A, Weil MH, Tang W, Pernat A, Bisera J. Optimizing timing of ventricular defibrillation. Crit Care Med 2001;29:2360–5. 24. 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. 25. Brown CG, Dzwonczyk R. Signal analysis of the human electrocardiogram during ventricular fibrillation: frequency and amplitude parameters as predictors of successful countershock. Ann Emerg Med 1996;27:184–8. 26. Neurauter A, Eftestol T, Kramer-Johansen J, et al. Prediction of countershock success using single features from multiple ventricular fibrillation frequency bands and feature combinations using neural networks. Resuscitation 2007;73:253–63. 27. Spohr F, Arntz HR, Bluhmki E, et al. International multicentre trial protocol to assess the efficacy and safety of tenecteplase during cardiopulmonary resuscitation in patients with out-of-hospital cardiac arrest: the Thrombolysis in Cardiac Arrest (TROICA) Study. Eur J Clin Invest 2005;35:315–23. 28. Boettiger BW. Thrombolysis in Cardiac Arrest (TROICA). Available at: http://www.escardio.org/knowledge/congresses/CongressReports/hotlinesandctus/ 709011 Boettiger.htm [accessed 19.06.08]. 29. Arntz HR, Wenzel V, Dissmann R, Marschalk A, Breckwoldt J, Muller D. Outof-hospital thrombolysis during cardiopulmonary resuscitation in patients with high likelihood of ST-elevation myocardial infarction. Resuscitation 2008;76:180–4. 30. Sunde K, Wik L, Steen PA. Quality of mechanical, manual standard and active compression-decompression CPR on the arrest site and during transport in a manikin model. Resuscitation 1997;34:235–42. 31. Olasveengen TM, Wik L, Steen PA. Quality of cardiopulmonary resuscitation before and during transport in out-of-hospital cardiac arrest. Resuscitation 2008;76:185–90.
T.M. Olasveengen et al. / Resuscitation 80 (2009) 412–417 32. Wagner H, Van der Pals J, Olasson HR, Gotberg M, Harnek J, Olivecrona G. Mechanical chest compression devices can save lives in the cath lab. Resuscitation 2008;77. S112 (AS-031). 33. ECC Committee, Subcommittees and Task Forces of the American Heart Association. Circulation 2005;112:IV-12.
417
34. Neurauter A, Eftestol T, Kramer-Johansen J, et al. Prediction of countershock success using single features from multiple ventricular fibrillation frequency bands and feature combinations using neural networks. Resuscitation 2007;73:253–63.
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical Paper
Acute ischemic heart disease alters ventricular fibrillation waveform characteristics in out-of hospital cardiac arrest夽 Theresa M. Olasveengen a,∗ , Trygve Eftestøl b , Kenneth Gundersen b , Lars Wik c , Kjetil Sunde a a
Institute for Experimental Medical Research and Department of Anaesthesiology, Ulleval University Hospital, N-0407 Oslo, Norway Department of Electrical and Computer Engineering, Faculty of Science and Technology, University of Stavanger, Stavanger, Norway c The National Competence Centre for Emergency Medicine and Institute for Experimental Medical Research, Ulleval University Hospital, N-0407 Oslo, Norway b
a r t i c l e
i n f o
Article history: Received 20 August 2008 Received in revised form 12 January 2009 Accepted 19 January 2009 Keywords: Advanced life support Cardiac arrest Electrocardiography Heart failure Ischemia Myocardial infarction Out-of-hospital CPR Ventricular fibrillation
a b s t r a c t Background: Although ventricular fibrillation waveform characteristics (VFWC) correlate with coronary perfusion pressure and may predict defibrillation outcome, recent animal data indicate that these waveform characteristics are altered in both acute myocardial infarction (AMI) and chronic coronary heart disease (CHD). We wanted to confirm these recent animal data in humans and explore the possibility for such characteristics to identify acute ischemia during cardiac arrest. Methods: Data from all adult patients admitted to hospital after out-of-hospital VF cardiac arrest in Oslo between May 2003 and July 2007 were prospectively collected. Patients were categorized into one of four pre-defined etiologic groups: patients with AMI (AMI only), patients with AMI and CHD (AMI and CHD), patients with previous CHD without evidence for a new AMI (CHD only), and patients with primary arrhythmia (PA). VFWC were analyzed from prehospital ECG tracings, and the different etiologic groups compared using ANOVA. Results: One-hundred-and-one patients with ECG recordings usable for VF analysis could confidently be categorized; 16 with AMI only, 34 with AMI and CHD, 41 with CHD only and 10 with PA. The two VFWC median slope (MS) and amplitude spectral area (AMSA) were significantly depressed in patients with AMI only compared to both PA (MS p = 0.008, AMSA p = 0.035) and CHD only patients (MS p = 0.008, AMSA p = 0.006). Conclusions: AMI patients have depressed MS and AMSA compared to patients without AMI during VF cardiac arrest. VFWC might be helpful in identifying patients with AMI during cardiac arrest, but prospective clinical studies are warranted to assess its feasibility and clinical benefit. © 2009 Elsevier Ireland Ltd. All rights reserved.
Introduction Approximately 30–40% of all out-of-hospital cardiac arrest (OHCA) patients are found with initial ventricular fibrillation (VF),1–3 and typically ∼20% of these patients survive to hospital discharge.3,4 It can be expected that 60–80% of the OHCA patients with cardiac aetiology have underlying acute or chronic ischemic heart disease.5,6 When angiography was performed on consecutive patients admitted to our hospital after OHCA with presumed cardiac aetiology ∼50% had acute myocardial infarction (AMI).7 There is wide consensus that early revascularization with percutaneous coronary intervention (PCI) or fibrinolytic therapy will
夽 A Spanish translated version of the summary of this article appears as Appendix in the final online version at doi:10.1016/j.resuscitation.2009.01.012. ∗ Corresponding author at: Institute for Experimental Medical Research, Ullevål University Hospital, University of Oslo, Division Ullevål University, Hospital, N-0407 Oslo, Norway. Tel.: +47 23016837/41419930; fax: +47 23016799. E-mail address: [email protected] (T.M. Olasveengen). 0300-9572/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.resuscitation.2009.01.012
decrease mortality for patients with ST-segment elevation myocardial infarction (STEMI),8,9 and that time from ischemic event to reperfusion is crucial.10,11 The same strategy has been advocated for cardiac arrest patients after return of spontaneous circulation (ROSC) when AMI is suspected.7,12–14 However, only approximately half of cardiac arrest patients with initial VF and presumed cardiac aetiology can be expected to achieve sustained ROSC during advanced life support (ALS),15 and prolonged resuscitation efforts are often ultimately terminated in the field with little else to offer these patients. Is has recently been demonstrated that coronary angiography and PCI is possible during continuous mechanical chest compressions,16,17 and may be a corrective treatment for a subgroup of cardiac arrest patients where ROSC is not achieved with traditional ALS. Patients with AMI and shock resistant VF cardiac arrest, but with “hearts and brains too good to die”, could be transported with mechanical CPR directly to angiography where underlying occlusions could be treated. Such a strategy could shorten the all important time to reperfusion and perhaps save lives of patients that would otherwise be terminated in the field. A major
T.M. Olasveengen et al. / Resuscitation 80 (2009) 412–417
challenge is early identification of patients who would benefit from this aggressive treatment. VF waveform analysis correlates with coronary perfusion pressure18 and has been described as a method to predict when a VF is likely to be shocked into a pulse generating rhythm, thereby reducing pauses in chest compressions and potential myocardial damage by limiting redundant shocks.19,20 It was recently shown that animal VF waveform characteristics appear altered after acute myocardial infarction (AMI)21 and chronic heart failure,22 and it may therefore be questionable to extrapolate earlier animal data where cardiac arrest has been induced electrically on structurally normal hearts. We are unaware of any studies on the effects of ischemic heart disease on human VF waveform characteristics. We hypothesize that VF waveform characteristics in patients with cardiac arrest and AMI and/or chronic coronary heart disease (CHD) will be different compared to patients with arrhythmic VF-arrest, and that VF analysis may contribute to identify ischemic conditions with indications for reperfusion therapy during cardiac arrest.
413
impedance changes. Annotations were made while reviewing available clinical information from the Utstein forms and ambulance records. Total time without spontaneous circulation (CPR time), time without chest compressions divided by CPR time (Hands-off ratio), compression rate and the actual number of compressions and ventilations per minute were calculated for each episode. Patient classification Patients were classified into the following four etiologic categories: (1) AMI without any previous symptoms or treatment for CHD (AMI only), (2) AMI in patients with previous symptoms or treatment for CHD (AMI and CHD), (3) Patients with previous symptoms or treatment for CHD without AMI (CHD only), (4) Primary arrhythmia without any previous symptoms or treatment for CHD (PA),
Material and methods Study design and recruitment All patients older than 18 years admitted to hospital after non-traumatic VF OHCA between May 2003 and July 2007 were retrospectively studied. Due to an ongoing randomized study of the effect of intravenous access and drugs in the Oslo Emergency Medical Service (EMS), some of our included patients are also included in this study registered at clinicaltri.gov (NCT00121524). Approximately half of the patients will therefore be expected to have received intravenous drugs sometime during the resuscitation effort. As we have used the first seconds of the ECG registration in our analysis (see VF characteristics below), randomization to the following protocols including or excluding intravenous drugs should be of little importance. The EMS system in Oslo is a one-tiered centralized community run system for a population of 540 000. In weekdays between 7:30 AM and 10 PM a physician-manned ambulance staffed by two paramedics and an anaesthesiologist functions on the same level as the regular paramedic staffed ambulances.
In addition, we registered whether the patients had been diagnosed with chronic heart failure prior to the arrest (typical clinical signs, use of heart failure medication and ejection fraction <50%, if available). The following information was considered when evaluating aetiology and degree of ischemia (from highest to lowest importance): autopsy reports, coronary intervention reports, previous medical history, biochemistry (including troponin release), decision to install an Implantable Cardioverter Defibrillator (ICD), and ECG findings. The AMI diagnosis was only used if there were clear angiographic or autopsy descriptions, or if there was a substantial and increasing Troponin release in the following first three blood samples. Small troponin release alone (<0.5 mcg/l) was not classified as AMI. Cases where the patient record did not contain sufficient information to determine the causes of cardiac arrest were excluded. The classifications were done independently by two physicians with extensive experience in prehospital ECG annotations/rhythm interpretation (TMO and KS), and any eventual differences in opinion were discussed until consensus was reached. Importantly, this classification was done independently and at a different location than the VF waveform calculations (TE and KG).
Data collection VF characteristics Utstein forms are routinely filled out by ambulance personnel after every cardiac arrest and submitted to the study supervisor along with a copy of the ambulance run sheet. Automated, computer based time records from the dispatch centre supplement ambulance run sheets with regard to response times. For all admitted patients, additional hospital records are obtained from the respective receiving hospitals. Information from Utstein forms, ambulance run sheets, dispatch and hospital records are linked together with continuous ECG tracings as described below. Equipment and data processing Standard LIFEPAK 12 defibrillators (Physio-Control, a Division of Medtronic, Redmond, WA, USA) were used, which routinely measure transthoracic impedance by applying a near constant sinusoidal current across the standard defibrillation pads. After a CPR effort the ECGs with transthoracic impedance signals were normally transferred to a local server at The National Competence Centre for Emergency Medicine (Ulleval University Hospital, Oslo, Norway), and data from each case were viewed and annotated using a CODE-STATTM 7.0 (Physio-Control, Redmond, WA, USA) for detection of chest compressions and ventilations from transthoracic
The data files for the included patients were acquired from the CODE-STAT system using a proprietary software program provided by Physio-Control for extracting the data. The extracted data files include defibrillator log data of important events, sampled measurements of electrocardiograms (ECG) and impedance and manual rhythm annotations. The sample rates of the ECG and impedance channels are 125 Hz and 61 Hz respectively. The log data, ECG, impedance data and rhythm annotations were integrated in MATLAB (The Mathworks Inc.) where further signal analysis was performed. All initial VF periods without compressions were identified. These chest compression pauses were generally pauses for on-scene initial rhythm assessment. The ECG tracings from these periods were extracted, and six different VF characteristics were computed for the last 125 samples of all identified tracings (corresponding to 1 s duration of the analysis window). Five previously studied VF characteristics based on analysis of spectral characterization were chosen. Prior to calculating the VF characteristics themselves the spectral representation of the signal segment to be analyzed is represented by its fast Fourier transform giving the amplitude and corresponding frequency of each of the signal’s frequency compo-
414
T.M. Olasveengen et al. / Resuscitation 80 (2009) 412–417
nents. The VF characteristics are derived from the signal’s spectral representation and were as follows: 1. The amplitude spectrum relationship (AMSA) is computed as the sum of weighted amplitude values weighted by frequency value, thus emphasizing high frequency content.23,24 2. Centroid frequency (CF) characteristics is computed as the frequency corresponding to the point of mass in the spectrum.25 Peak power frequency (PPF) corresponds to the frequency of the maximum spectral amplitude.25 The spectral flatness measure (SFM) is computed as the ratio of the geometric mean and the arithmetic mean of the spectrum.20 3. The signal energy (ENRG) is computed as the area under the spectrum.20 4. The median slope (MS) is computed as the median value of amplitude differences in the analysis window, thus also emphasizing the high frequency content.26 AMSA and MS combine information about amplitude and frequency so that high frequency and amplitude VF (coarse VF) give high AMSA and MS values. Statistical analysis Statistical calculations were performed using a spreadsheet program (Excel 2002, Microsoft Corp, Redmond, WA, USA) or a statistical software package (SPSS 14.0, SPSS Inc., Chicago, IL, USA). Natural logarithm was computed for MS, AMSA, CF and ENRG values before statistical analysis to change the distribution of the parameters to be approximately Gaussian. Comparisons between the different etiologic groups were done with ANOVA and independent samples t-test with Bonferroni corrections. p-values less than 0.05 were considered significant. Results From May 2003 to July 2007 the Oslo EMS responded to 1867 cardiac arrests. ALS was attempted in 931 cases, whereof 312 (34%) were found in VF. Among the VF patients, 205 (66%) were admitted to hospital. One-hundred and seven (52%) of the 205 cases eligible for inclusion had usable ECG recordings with impedance signals. Six additional cases had to be excluded as they could not confidently be classified into one of the four defined etiologic categories from available patient information, leaving 101 cases (49%) included for further analysis. General Utstein data with CPR quality parameters are presented in Table 1. There were no significant differences in Utstein characteristics between included patients and the patient group as a whole (Table 1). Among the 101 included patients, 37 had previously been diagnosed and treated for chronic heart failure while 54 had not exhibited clinical signs of chronic heart failure prior to their cardiac arrest (10 unknown). Fifty percent of the patients in the AMI + CHD group and 61% in the CHD group had confirmed chronic heart failure prior to arrest (none in the two other groups). With ANOVA two of the six VF waveforms characteristics varied significantly between the etiologic groups; MS (p = 0.003) and AMSA (p = 0.005) (Table 2). With post hoc Bonferroni corrected independent t-tests both MS and AMSA were significantly depressed for the AMI only group compared to both PA only (p = 0.008 and 0.035, respectively) and CHD only groups (p = 0.008 and 0.006, respectively). The overall trend for the VF waveform characteristics were values increasing from the AMI only → AMI and CHD → CHD only → PA (Table 2). Compared to all patients without AMI, all patients with AMI had significantly depressed MS (4.3 vs. 5.5, p = 0.002), AMSA (7.8
Table 1 Utstein data with CPR quality parameters for out-of-hospital cardiac arrest patients who present with initial ventricular fibrillation and are subsequently admitted to hospital. All VF cardiac arrests admitted to hospital (n = 205)
Included VF cardiac arrests admitted to hospital (n = 101)
Female Age (years) Cardiac aetiology
52 (25) 64 ± 16 186 (91)
22 (22) 66 ± 16 96 (95)
Location of arrest Home Work Public place Other
73 (36) 12 (6) 102 (50) 18 (9)
31 (31) 4 (4) 58 (57) 7 (7)
Cardiac arrest witnessed by Bystanders AED-trained bystanders Ambulance personnel Not witnessed
151 (74) 7 (3) 29 (14) 18 (9)
84 (83) 2 (2) 3 (3) 12 (12)
Bystander CPR Endotracheal intubation Response time (min) Time from arrest to ALS (min) CPR quality Hands-off ratio Compression rate Compressions (min−1 ) Ventilations (min−1 )
132 (64) 152 (74) 7 (5.8) 8 (7.9)
70 (69) 86 (85) 7 (6.8) 8 (7.10)
N/A N/A N/A N/A
0.24 ± 0.14 118 ± 11 88 ± 17 12 ± 4
Transported during ongoing CPR Admitted to ICU Discharged alive
47 (23)
20 (20)
160 (78) 95 (46)
90 (89) 50 (50)
CPC score 1 2 3 4
a
b
79 8 5 0
40 5 4 0
Values given as number of cases with percentage except from age and CPR quality parameters given as mean with standard deviation and response times given as median with 95% confidence intervals. a CPC classification not available for three patients. b CPC classification not available for one patient.
vs. 13.3, p = 0.003) and ENRG (893 vs. 1210, p = 0.025). Initial MS values in patients with and without AMI are shown in Figure 1. The values for the remaining VF waveform characteristics, CF, PPF and FLAT, were also lower in the patients with AMI compared to
Figure 1. Box-plot of ventricular fibrillation (VF) initial median slope (MS) for all patients with acute myocardial infarction (AMI) and all patients without AMI. pValue obtained from an independent t-test.
T.M. Olasveengen et al. / Resuscitation 80 (2009) 412–417
415
Table 2 Description of VF waveform characteristics for the different etiologic groups. VF waveform characteristic
AMI only (n = 16)
AMI + CHD (n = 34)
CHD only (n = 41)
PA (n = 10)
ANOVA (p-value)
Median slope (MS) Amplitude spectral area (AMSA) Energy (ENRG) Centroid frequency (CF) Peak power frequency (PPF) Spectral flatness (FLAT)
3.7 (2.4, 5.5) 6.2 (2.7, 11.7) 781 (339, 1624) 5.1 (4.1, 6.4) 4.3 (2.5, 5.4) 0.086 (0.065, 0.152)
4.6 (3.7, 5.5) 9.4 (4.6, 15.3) 1025 (662, 1466) 5.1 (4.8, 6.3) 4.4 (3.7, 5.9) 0.112 (0.101, 0.126)
4.9 (4.3, 6.1) 11.7 (7.1, 19.3) 1206 (928, 1591) 6.0 (5.0, 6.8) 4.9 (3.7, 5.9) 0.106 (0.078, 0.136)
6.1 (3.7, 9.8) 15.7 (4.5, 50.2) 1840 (551, 2984) 6.7 (3.6, 7.9) 5.8 (3.7, 7.4) 0.144 (0.055, 0.192)
0.003 0.005 0.083 0.505 0.526 0.879
VF waveform characteristics are given as median values (95% confidence intervals). The etiologic groups were compared using ANOVA. Natural logarithm was computed for MS, AMSA, ENRG and CF values before statistical analysis to improve normal distribution.
Table 3 VF waveform characteristics in patients with and without acute myocardial infarction. VF waveform characteristic
Patients with AMI (n = 50)
Patients without AMI (n = 51)
p-Value
Median slope (MS) Amplitude spectral area (AMSA) Energy (ENRG) Centroid frequency (CF) Peak power frequency (PPF) Spectral flatness (FLAT)
4.3 (3.1, 5.5) 7.8 (4.4, 12.0) 893 (703, 1100) 5.1 (4.6, 6.3) 4.4 (4.2, 5.1) 0.107 (0.080, 0.126)
5.5 (4.3, 6.7) 13.3 (7.1, 19.3) 1210 (833, 1798) 6.1 (5.0, 6.9) 4.9 (3.7, 5.9) 0.111 (0.076, 0.152)
0.002 0.003 0.025 0.284 0.520 0.616
Values in patients with and without acute myocardial infarction (AMI) were compared using independent t-tests. Natural logarithm was computed for MS, AMSA, ENRG and CF values before statistical analysis in order to approach a normal distribution.
the patients without AMI—but the differences were not significant (Table 3). Similarly, patients with chronic heart failure tended to have slightly depressed VF waveforms characteristics compared to those without chronic heart failure, although there were no statistically significant differences: AMSA 9.6 vs. 11.7 (p = 0.460), ENRG 1083 vs. 1100 (p = 0.305), CF 5.6 vs. 5.8 (p = 0.850), PPF 4.9 vs. 4.7 (p = 0.677), FLAT 0.12 vs. 0.10 (p = 0.549), respectively. Median MS was 4.9 in both groups. Discussion The present human data confirm recent animal data21,22 showing that ischemic heart disease alters VF waveform characteristics. Acute ischemic events seem to have a larger impact on VF characteristics than chronic ischemic disease and heart failure. VF waveform analysis could therefore potentially be useful in identifying patients with underlying AMI not achieving ROSC with traditional ALS, and this additional information could influence the ALS providers with regard to further treatment strategy. Early revascularization could be an option in patients with refractory VF and low VF waveform characteristics, selecting these patients for either on-scene thrombolytic therapy,27 or transport to the nearest angiographic laboratory for acute PCI utilizing a mechanical chest compression devise.16 Although thrombolytic treatment was not found to increase survival in a large randomized clinical trial of all patients with cardiac arrest with presumed cardiac cause (TROICA),27,28 the subgroup with suspected AMI could still benefit from such treatment.29 Early PCI during mechanical CPR is another strategy that might potentially increase survival for AMI patients with shock resistant VF.16 In order for PCI to be feasible before ROSC, the time from arrest to PCI facility arrival must be relatively short with a mechanical chest compression devise being used to secure adequate perfusion.30,31 PCI during CPR is costly and resource consuming, and such a treatment protocol would be dependent on good selection criteria to identify patients expected to benefit from such treatment.17,32 We tested six different VF waveform characteristics known to have varying shock prediction abilities. Although we find significantly depressed VF waveform characteristics (MS, AMSA and
ENRG) in patients with AMI compared to patients without AMI, the box plot in Figure 1 clearly illustrates that even the most promising characteristic (MS) does not provide a well defined cut-off value for clinical use. These characteristics would in their current form not give conclusive information on the presence of an AMI, but could be combined with other factors such as previous history, symptoms prior to arrest, age, anoxia time to aid ALS providers in their decision-making. Exploring other waveform characteristics, or modifying the characteristics already described, might yield a more clinically useful “AMI indicator” for cardiac arrest patients with shock resistant VF. Increasing research efforts to improve our knowledge and help identify the underlying mechanisms of cardiac arrest has been advocated, and is hoped to provide specific, corrective therapy, treating the underlying cause of arrest.33 The best shock predictors are generally found to be AMSA and MS with sensitivities ∼95% and specificities ∼60% when performed retrospectively.23,24,34 The usefulness of these VF characteristics are still being debated, as they have never been proven useful in a prospective clinical setting. The usefulness of VF analysis during cardiac arrest, both in shock prediction in general and in other applications such as AMI detection, needs to be investigated prospectively. The most important limitation to this study is the patient selection, as only patients admitted to hospital could be classified into our four etiologic groups. Ideally, all patients with initial VF should have been included and classified according to PCI reports, biochemical markers, ECG findings or autopsy results. As only a small fraction of cardiac arrest deaths are autopsied in Oslo, and virtually no patients dying before hospital admission, such information is not available. Exploring relationships between VF characteristics and structural abnormalities in the heart should be explored further, and in order to test the usefulness of VF analysis in OHCA, a prospective randomized trial is warranted. Another limitation for our data is that they are primarily gathered for another ongoing randomized interventional study on the effects of IV drugs during resuscitation which entails that two different treatment strategies are given to these patients. Any potential effect on VF characteristics from IV drugs such as epinephrine or amiodarone could be a confounding factor, but there is no rea-
416
T.M. Olasveengen et al. / Resuscitation 80 (2009) 412–417
son to believe the randomization to be unevenly distributed among the etiologic groups. Although this provides us with prospectively collected data, the present study design is retrospective. Conclusion Patients with ventricular fibrillation after acute myocardial infarction (AMI) have altered waveform characteristics compared to patients without AMI. Further exploration of such characteristics could yield a tool that might be helpful for identifying patients with AMI with the potential of targeted revascularization therapy during ALS. Prospective clinical studies are warranted to assess its feasibility and clinical benefit. Conflict of interest Olasveengen has received honoraria from Medtronic (Oslo, Norway) and research support from Laerdal Medical Corporation (Stavanger, Norway). Eftestol is the project leader and Gundersen a participant of a project granted by the Research Council of Norway where Laerdal Medical amongst others has contributed with data. Wik is on a Medical Advisory Board for Physio-Control, has in the past consulted for Laerdal and Jolife, and is the principle investigator for a multi-centre mechanical chest compression device study sponsored by Zoll. Sunde has no conflicts of interest to declare. Acknowledgements We thank all physicians and paramedics working in the Oslo EMS Service, as well as Martin Samdal for invaluable help in obtaining patient records, and especially Petter Andreas Steen for discussions and critique. The study was supported by grants from Eastern Norway Regional Health Authority, Ulleval University Hospital, Norwegian Air Ambulance Foundation, Laerdal Foundation for Acute Medicine, and Anders Jahres Fund. Software necessary for analysing data from the LIFEPAK 12 defibrillators was provided by PhysioControl (Redmond, WA, USA). Financial support: The study was supported by grants from Eastern Norway Regional Health Authority, Ulleval University Hospital, Laerdal Foundation for Acute Medicine, Anders Jahres Fund, The Research Council of Norway. The necessary software for analyzing data from the LIFEPAK 12 defibrillators was provided by Physio-Control, Redmond, WA, USA. Author contributions: Olasveengen and Eftestøl have full access to generated data and take full responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Olasveengen, Eftestøl, and Sunde. Acquisition of data: Olasveengen and Wik. Analysis and interpretation of data: Olasveengen, Eftestøl, and Gundersen. Drafting of the manuscript: Olasveengen, Eftestøl, and Sunde. Critical revision of the manuscript for important intellectual content: Olasveengen, Eftestøl, Gundersen, Wik, and Sunde. Statistical expertise: Olasveengen, Eftestøl, and Gundersen. Obtained funding: Sunde, Wik, and Eftestøl. Study supervision: Sunde. References 1. Cobb LA, Fahrenbruch CE, Olsufka M, Copass MK. Changing incidence of out-ofhospital ventricular fibrillation, 1980–2000. JAMA 2002;288:3008–13. 2. Herlitz J, Engdahl J, Svensson L, Young M, Angquist KA, Holmberg S. Decrease in the occurrence of ventricular fibrillation as the initially observed arrhythmia after out-of-hospital cardiac arrest during 11 years in Sweden. Resuscitation 2004;60:283–90. 3. Atwood C, Eisenberg MS, Herlitz J, Rea TD. Incidence of EMS-treated out-ofhospital cardiac arrest in Europe. Resuscitation 2005;67:75–80. 4. Rea TD, Eisenberg MS, Sinibaldi G, White RD. Incidence of EMS-treated out-ofhospital cardiac arrest in the United States. Resuscitation 2004;63:17–24.
5. Silfvast T. Cause of death in unsuccessful prehospital resuscitation. J Int Med 1991;229:331–5. 6. Sunde K, Pytte M, Jacobsen D, et al. Implementation of a standardised treatment protocol for post resuscitation care after out-of-hospital cardiac arrest. Resuscitation 2007;73:29–39. 7. Spaulding CM, Joly LM, Rosenberg A, et al. Immediate coronary angiography in survivors of out-of-hospital cardiac arrest. N Engl J Med 1997;336:1629–33. 8. Fibrinolytic Therapy Trialists’ (FTT) Collaborative Group. Indications for fibrinolytic therapy in suspected acute myocardial infarction: collaborative overview of early mortality and major morbidity results from all randomised trials of more than 1000 patients. Lancet 1994;343:311–22. 9. Keeley EC, Boura JA, Grines CL. Primary angioplasty versus intravenous thrombolytic therapy for acute myocardial infarction: a quantitative review of 23 randomised trials. Lancet 2003;361:13–20. 10. Antman EM, Anbe DT, Armstrong PW, et al. ACC/AHA guidelines for the management of patients with ST-elevation myocardial infarction: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Revise the 1999 Guidelines for the Management of Patients with Acute Myocardial Infarction). Circulation 2004;110:E82–292. 11. Van de WF, Ardissino D, Betriu A, et al. Management of acute myocardial infarction in patients presenting with ST-segment elevation. The Task Force on the Management of Acute Myocardial Infarction of the European Society of Cardiology. Eur Heart J 2003;24:28–66. 12. Bendz B, Eritsland J, Nakstad AR, et al. Long-term prognosis after out-of-hospital cardiac arrest and primary percutaneous coronary intervention. Resuscitation 2004;63:49–53. 13. Keelan PC, Bunch TJ, White RD, Packer DL, Holmes Jr DR. Early direct coronary angioplasty in survivors of out-of-hospital cardiac arrest. Am J Cardiol 2003;91, 1461-3, A. 14. Quintero-Moran B, Moreno R, Villarreal S, et al. Percutaneous coronary intervention for cardiac arrest secondary to ST-elevation acute myocardial infarction. Influence of Immediate Paramedical/Medical Assistance on Clinical Outcome. J Invasive Cardiol 2006;18:269–72. 15. Sunde K, Eftestol T, Askenberg C, Steen PA. Quality assessment of defribrillation and advanced life support using data from the medical control module of the defibrillator. Resuscitation 1999;41:237–47. 16. Grogaard HK, Wik L, Eriksen M, Brekke M, Sunde K. Continuous mechanical chest compressions during cardiac arrest to facilitate restoration of coronary circulation with percutaneous coronary intervention. J Am Coll Cardiol 2007;50:1093–4. 17. Larsen AI, Hjornevik AS, Ellingsen CL, Nilsen DW. Cardiac arrest with continuous mechanical chest compression during percutaneous coronary intervention: a report on the use of the LUCAS device. Resuscitation 2007;75:454–9. 18. Noc M, Weil MH, Tang W, Sun S, Pernat A, Bisera J. Electrocardiographic prediction of the success of cardiac resuscitation. Crit Care Med 1999;27:708–14. 19. Xie J, Weil MH, Sun S, et al. High-energy defibrillation increases the severity of postresuscitation myocardial dysfunction. Circulation 1997;96:683–8. 20. Eftestol T, Sunde K, Ole AS, Husoy JH, Steen PA. Predicting outcome of defibrillation by spectral characterization and nonparametric classification of ventricular fibrillation in patients with out-of-hospital cardiac arrest. Circulation 2000;102:1523–9. 21. Indik JH, Donnerstein RL, Berg RA, Hilwig RW, Berg MD, Kern KB. Ventricular fibrillation frequency characteristics are altered in acute myocardial infarction. Crit Care Med 2007;35:1133–8. 22. Indik JH, Donnerstein RL, Kern KB, Goldman S, Gaballa MA, Berg RA. Ventricular fibrillation waveform characteristics are different in ischemic heart failure compared with structurally normal hearts. Resuscitation 2006;69:471–7. 23. Marn-Pernat A, Weil MH, Tang W, Pernat A, Bisera J. Optimizing timing of ventricular defibrillation. Crit Care Med 2001;29:2360–5. 24. 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. 25. Brown CG, Dzwonczyk R. Signal analysis of the human electrocardiogram during ventricular fibrillation: frequency and amplitude parameters as predictors of successful countershock. Ann Emerg Med 1996;27:184–8. 26. Neurauter A, Eftestol T, Kramer-Johansen J, et al. Prediction of countershock success using single features from multiple ventricular fibrillation frequency bands and feature combinations using neural networks. Resuscitation 2007;73:253–63. 27. Spohr F, Arntz HR, Bluhmki E, et al. International multicentre trial protocol to assess the efficacy and safety of tenecteplase during cardiopulmonary resuscitation in patients with out-of-hospital cardiac arrest: the Thrombolysis in Cardiac Arrest (TROICA) Study. Eur J Clin Invest 2005;35:315–23. 28. Boettiger BW. Thrombolysis in Cardiac Arrest (TROICA). Available at: http://www.escardio.org/knowledge/congresses/CongressReports/hotlinesandctus/ 709011 Boettiger.htm [accessed 19.06.08]. 29. Arntz HR, Wenzel V, Dissmann R, Marschalk A, Breckwoldt J, Muller D. Outof-hospital thrombolysis during cardiopulmonary resuscitation in patients with high likelihood of ST-elevation myocardial infarction. Resuscitation 2008;76:180–4. 30. Sunde K, Wik L, Steen PA. Quality of mechanical, manual standard and active compression-decompression CPR on the arrest site and during transport in a manikin model. Resuscitation 1997;34:235–42. 31. Olasveengen TM, Wik L, Steen PA. Quality of cardiopulmonary resuscitation before and during transport in out-of-hospital cardiac arrest. Resuscitation 2008;76:185–90.
T.M. Olasveengen et al. / Resuscitation 80 (2009) 412–417 32. Wagner H, Van der Pals J, Olasson HR, Gotberg M, Harnek J, Olivecrona G. Mechanical chest compression devices can save lives in the cath lab. Resuscitation 2008;77. S112 (AS-031). 33. ECC Committee, Subcommittees and Task Forces of the American Heart Association. Circulation 2005;112:IV-12.
417
34. Neurauter A, Eftestol T, Kramer-Johansen J, et al. Prediction of countershock success using single features from multiple ventricular fibrillation frequency bands and feature combinations using neural networks. Resuscitation 2007;73:253–63.