Priming the Pump—Can Delaying Defibrillation Improve Survival After Sudden Cardiac Death?

Few medical phenomena engage the popular and medical imaginations as completely as resuscitation of the patient with out-of-hospital cardiac arrest. Application of defibrillator paddles to the chest wall, administering an electrical countershock, and restoring cardiac rhythm and circulation is depicted in films and on television as a dramatic, often lifesaving event.

The medical reality, however, is that survival after out-of-hospital cardiac arrest is uncommon. Despite decades of research and advances in resuscitation, less than 5% of patients survive out-of-hospital cardiac arrest in the United States.^{1 - 2} These outcomes have prompted the suggestion that new approaches for treatment of cardiac arrest due to ventricular fibrillation should be developed.³ In this issue of THE JOURNAL, Wik and colleagues⁴ report the results of a prospective, randomized comparison of manual cardiopulmonary resuscitation (CPR) prior to defibrillation vs traditional immediate defibrillation in patients with out-of-hospital cardiac arrest.

A substantial proportion of cases of out-of-hospital cardiac arrest are due to ventricular fibrillation.^{5 - 6} Ventricular fibrillation is characterized by a chaotic, irregular waveform on the electrocardiogram and results in a quivering heart that is unable to pump blood. The echocardiographic waveform loses amplitude with time, culminating in complete loss of cardiac electrical activity (asystole).⁷ At the cellular level, ongoing ventricular fibrillation consumes intracellular adenosine triphosphate, increases intracellular calcification, and causes other metabolic disturbances that ultimately result in myocardial cell death.⁸ Clinically, ventricular fibrillation manifests as cardiovascular collapse—with loss of pulse and spontaneous breathing.⁹

The definitive treatment for ventricular fibrillation is electrical countershock. The intervals from collapse to initiation of CPR and from collapse to defibrillation predict the probability of successful defibrillation and patient survival, although the window of opportunity is cruelly brief. Defibrillation performed within 3 minutes of collapse yields survival to hospital discharge in 70% of cases of out-of-hospital cardiac arrest due to ventricular fibrillation.¹⁰ Initiation of manual CPR improves the probability of survival for any given duration of ventricular fibrillation¹¹ ; however, survival is essentially zero after more than 10 minutes of uninterrupted fibrillation.

Animal studies first suggested that a period of manual CPR prior to electrical defibrillation might increase the probability of successful conversion to a perfusing cardiac electrical rhythm rather than asystole in subjects with ventricular fibrillation of longer duration.¹² The model developed by Niemann et al¹² suggested that the benefits of "preparing" the fibrillating myocardium with CPR and high-dose epinephrine prior to electrical countershock seemed to outweigh the delay of defibrillation.

Cobb et al¹³ used the findings of Niemann et al¹² as one rationale for the implementation in 1994 of a protocol of CPR prior to defibrillation for out-of-hospital cardiac arrest in Seattle, Wash. Another rationale was the observation that widespread first responder defibrillation in Seattle did not result in improvement of survival after out-of-hospital ventricular fibrillation despite a reduction in the delay to defibrillation. The protocol did not attempt to differentiate between ventricular fibrillation of longer duration vs shorter duration nor was it intended to examine the efficacy of CPR first protocols. However, Cobb et al found that patients who experienced cardiac arrest and had emergency medical service dispatch-to-arrival response intervals of longer than 4 minutes experienced improved survival compared with historical controls. Equally important, the probability of survival was not adversely affected for patients whose response intervals were shorter than 4 minutes.

These observations are important for at least 3 reasons. First, the outcomes were consistent with prior animal investigations suggesting that preparing the myocardium with manual CPR enhanced the probability of reestablishing a perfusing cardiac rhythm and survival after defibrillation. Second, the benefit accrued largely to patients for whom emergency medical service personnel responded in 4 minutes or longer. Third, delaying defibrillation for 90 seconds to provide 150 chest compressions did not adversely affect the probability of survival for individuals whose response intervals were shorter than 4 minutes. The Seattle study, even though it involved a quasi-experimental, before and after design, provided an ethical basis for randomization in a prospective trial of CPR prior to defibrillation for out-of-hospital ventricular fibrillation.

The report by Wik et al from Oslo is the first of such prospective randomized trials. Patients received either standard care for ventricular fibrillation with immediate defibrillation or received CPR first with 3 minutes of basic CPR followed by defibrillation. Overall, there were no differences between the standard group and CPR first group in terms of return of spontaneous circulation, survival to hospital discharge, or 1-year survival. However, in the subgroup of patients with ambulance response intervals longer than 5 minutes, more patients who received CPR first than those who received standard defibrillation care had return of spontaneous circulation, survived to hospital discharge, and experienced survival to 1 year.

Interpretation of the findings reported by Wik et al requires acknowledgment of several limitations. First, the interval of interest in out-of-hospital cardiac arrest is the ischemic interval, which is the period bounded by patient collapse and the reestablishment of perfusion to the brain and heart.⁹ Response intervals are just one part of the ischemic interval and do not account for the time required to recognize the individual's collapse, time to call for emergency assistance, or time for the dispatcher handling the call to dispatch emergency medical service personnel. At the other end, the response interval does not include the time required for emergency medical service personnel to reach the patient after arrival on scene, to deploy the defibrillator, and to deliver the first countershock. Cardiac arrest researchers who use the response interval as a predictor variable hope to avoid systematic bias in the unaccounted for components of the ischemic interval. Second, Wik et al stratified patients and reported a subgroup analysis using a response interval of 5 minutes or shorter or longer than 5 minutes. This approach essentially neutralizes the randomization scheme and renders the study unable to address the question of interest robustly—if patients with response intervals of longer than five minutes were treated with a CPR first protocol and patients with response intervals of 5 minutes or shorter were immediately defibrillated, would overall survival improve? Third, Wik et al report their data in the Utstein style for uniform reporting of data on cardiac arrest, but did not include several data elements specified in the Utstein template (such as time of patient collapse, time called for assistance received by a dispatcher, time of responder arrival at patient).¹⁴ This deficiency limits the generalizability of their findings and makes it difficult to compare their findings with the research of other investigators in other venues.

These limitations do not detract from the potential clinical importance of the study by Wik et al. Their findings are remarkably similar to those reported by Cobb et al.¹³ These results may reasonably be taken as confirmation that in cases with longer ischemic intervals, CPR prior to defibrillation may produce better survival than traditional immediate defibrillation in out-of-hospital ventricular defibrillation. Further trials are needed in which these limitations are ameliorated; some desirable elements for future trials are immediately apparent.

The benefit of CPR first seems to accrue only among cases of longer duration ventricular fibrillation. Therefore, a method needs to be incorporated that permits rapid identification of shorter or longer duration ventricular fibrillation. As Wik et al note, characteristics of the initial ventricular fibrillation waveform may be helpful. Brown et al¹⁵ found that the amplitude and frequency of the ventricular fibrillation waveform could be combined into a measure (median frequency of the waveform power spectrum) that was unaffected by chest wall thickness or shape of the thorax and correlated with the duration of ventricular fibrillation. With continued advances in defibrillation technology, it may be feasible to use waveform analysis to identify patients with long duration vs those with short duration ventricular fibrillation, and thereby assign them to a treatment of CPR first or defibrillation first. A study design randomizing patients with long duration ventricular fibrillation to CPR first or immediate defibrillation strategies would also be scientifically desirable, and may be an important next step in research.

In addition, if more patients who experience long duration ventricular fibrillation survive to hospital discharge, the incidence and severity of neurological impairment and impaired health-related quality of life among out-of-hospital cardiac arrest survivors may increase. Although recommended in the Utstein style and long used in cardiac arrest investigations, the Glasgow-Pittsburgh Outcome Categories distinguish only gross differences in patient outcome status (eg, category 2 indicates conscious with moderate disability; category 3, conscious with severe disability).¹⁶ Use of more sophisticated functional assessments and quality-of-life measures will be required to provide data to clarify this issue.

Improving survival after out-of-hospital cardiac arrest requires multiple strategies. Applying defibrillation earlier to a greater proportion of patients with ventricular fibrillation of short duration is the aim of the Public Access Defibrillation initiative,^{17 - 18} which is an ongoing clinical trial. The study by Wik et al suggests that protocols consisting of CPR prior to defibrillation offer an attractive complementary possibility for improving survival among patients with ventricular fibrillation of longer duration, and deserve further evaluation.