How many prehospital debrillation attempts should be performed in out-of-hospital cardiac arrest patients with shockable rhythm?

Background: Early debrillation is vital to improve outcomes after out-of-hospital cardiac arrest (OHCA) with shockable rhythm. Currently, there is no agreed consensus on the number of debrillation attempts before transfer to a hospital. This study aimed to evaluate the correlation between the number of debrillations on the prehospital return of spontaneous circulation (ROSC). Methods: A multicenter, prospective, observational registry-based study was conducted for OHCA in patients with presumed cardiac etiology that underwent prehospital debrillation between October 2015 and June 2017. The primary outcome was prehospital ROSC, and the secondary outcome was a good neurologic outcome at hospital discharge, dened as Cerebral Performance Category score 1 or 2. Results: Among 2,155 OHCA patients’ data, 178 patients with missing data were excluded, a total of 1,983 OHCA patients who received prehospital debrillation were included. The median age was 61 years and prehospital ROSC was observed in 738 patients (37.2%). The median time from arrest to rst debrillation was 10 (interquartile range: 7-15) minutes. The cumulative ROSC rates and good neurologic outcome from the initial debrillation to the sixth debrillation were 43%, 68%, 81%, 90%, 95%, 98% and 42%, 66%, 81%, 90%, 95%, 98%, respectively. After clinical characteristics adjustment and time to debrillation, the number of debrillations were independently associated with ROSC (odds ratio 0.81 95% CI 0.76-0.86) and good neurologic outcome (odds ratio 0.86 95% CI 0.80-0.91). Moreover, subgroup analysis results with patients that underwent the initial debrillation within 10 minutes from arrest were consistent (95% up to ve times). Conclusion: More than 95% of prehospital ROSC was achieved within ve times of debrillation in OHCA patients. This result provides a basis for the ideal number of debrillation attempts before transfer to hospital with the possibility of extracorporeal cardiopulmonary resuscitation in these refractory ventricular brillation patients.


Introduction
With advances in cardiac arrest resuscitation, the rate of survival discharge in patients with out-ofhospital cardiac arrest (OHCA) who receive cardiopulmonary resuscitation (CPR) have been reported ranging from 7.5-10.8% [1][2][3][4]. The rst monitored rhythm is ventricular brillation (VF) and pulseless ventricular tachycardia (pVT) in approximately 20% of cardiac arrests, but the incidence of VF/pVT can vary according to bystander CPR rates [5][6][7][8]. VF/pVT will also occur at some stage during resuscitation in about 25% of cardiac arrests with an initial documented rhythm of asystole or pulseless electrical activity [9].
Early de brillation, one of the chain of survival, plays an important role in improving cardiac arrest patients' survival from VF/Pvt [10]. During treatment of VF/pVT, with every minute that passes between collapse and de brillation, survival rates decrease by 7-10% if no CPR is provided, and 3-4% if bystander CPR is provided [11,12]. Guidelines for CPR have recommended a single-shock strategy without escalating the shock energy, without delay. If ROSC has not been achieved by the third shock, the adrenaline may increase the chance of successful de brillation with the next shock. However, the optimal number of prehospital de brillation attempts in OHCA patients with VF/pVT before transfer to the hospital is unknown. Since repetitive de brillation is required to treat VF refractory to de brillation, knowledge of the effective number of de brillations is important in the initial treatment process for VF. As recurrence of VF after rst de brillation is common, two or more de brillations are typically necessary [13]. One study reported that 48% of the cohort presented with re brillation within two minutes of the rst de brillation shock, and 74% received at least one shock for re brillation during prehospital care [13].
Patients with refractory VF can be candidates for rescue treatments such as extracorporeal CPR if they arrive at the hospital within a reasonable time.
This study aimed to investigate de brillation success rates based on the number of de brillation attempts on OHCA patients that received prehospital de brillation.

Study population
This retrospective review of multicenter, prospective, observational registry was conducted between October 2015 and June 2017, using data from the Korean Cardiac Arrest Research Consortium (KoCARC). KoCARC is a multi-institutional, nationwide collaborative research network of 62 participating institutions developed to investigate the various studies conducted in the eld of OHCA and to enhance collaborative efforts to carry out the studies. [14] The KoCARC registry was designed to include OHCA patients that had been transported to the participating emergency departments (ED) by emergency medical services (EMS) with resuscitation efforts and patients who had a medical etiology identi ed by an emergency physician.
The registry excluded OHCA patients with terminal illness documented by medical records, hospice care, pregnancy, and pre-documented "No Resuscitation" cards. Also excluded were those with clear nonmedical etiology including trauma, drowning, poisoning, burns, asphyxiation, or hanging. The quality control committee's quality assurance plan includes integrity checks for required elds, and built-in validation rule cross-checks for data elds. They also provide feedback to research coordinators and investigators regarding quality control processes through quarterly meetings. The KoCARC data collection protocol was reviewed and approved by the Institutional Review Board of each participating hospital. The KoCARC Registry was registered at Clinicaltrials.gov (Identi er NCT03222999). A detailed description of the registry has been presented elsewhere [14,15].

Study Design And Data Variables
In addition to the general registry described above, this study used a speci c design. Our study only included OHCA in patients with presumed cardiac etiology who underwent prehospital de brillation. Cardiac etiology was de ned as a case of receiving de brillation treatment at least once. The patients without information on de brillation were excluded. We also excluded patients without information on outcome.
Patient characteristics (age, gender), prehospital characteristics (prehospital initial rhythm, bystander CPR, witnessed by a bystander, automated external de brillator use and de brillation). Drug administration by EMS personnel, prehospital advanced airway, prehospital ROSC (de ned as the restoration of a palpable pulse ≥ 20 minutes), time interval (arrest time to initial de brillation, response time de ned as time from EMS call to scene arrival, scene time de ned as the time from scene arrival to scene departure, and transport time de ned as the time from scene departure to hospital arrival), and hospital outcomes from the KoCARC registry.
The de brillation strategies for VF/pVT in South Korea follow the current guidelines for CPR, including single shock and not escalating energy dose for subsequent shocks. However, the duration of eld resuscitation and de brillation attempts are varied by their medical directors [16]. EMS personnel were unable to abandon on-site resuscitation attempts unless OHCA patients showed obvious signs of death.
Overt signs of death are de ned as the presence of (1) head-body disconnection, (2) rot, or (3) postmortem rigidity. Therefore, all EMS-treated OHCA patients should be transferred to the ED.
The success of prehospital de brillation was de ned as the achievement of ROSC. The primary outcome of this study was prehospital ROSC rates. The secondary outcome was a good neurologic outcome at the point of hospital discharge, and was de ned as a Cerebral Performance Category (CPC) score of 1 (good cerebral performance) and 2 (moderate cerebral disability) [17]. Multivariable logistic regression analysis was used to con rm whether the number of de brillations were independently associated with outcomes after adjusting for clinical characteristics and time to initial de brillation.

Statistical analysis
Continuous variables were analyzed as mean ± standard deviation or median of the interquartile range (IQR), and categorical variables were analyzed as absolute or relative frequency. Student's t-test or Mann-Whitney u-test were used to compare continuous variables, and categorical variables were chi-square test or Fisher's exact test.
The baseline characteristics of the whole study population were analyzed. Multivariable logistic regression analysis was used to determine the predictors for prehospital ROSC and good neurologic outcome. Potential confounding factors that were adjusted for in the multivariable analyses included sex, age, prehospital initial rhythm, bystander CPR, witnessed by a bystander, number of de brillation attempts, drug administration by EMS personnel, prehospital advanced airway and time interval (arrest time, response time, scene time and transport time).
Cumulative incidence of prehospital ROSC and good neurologic outcome according to the number of de brillations were examined. Subgroup analysis of patients who underwent de brillation within the median time from arrest rst de brillation was also assessed to minimize the effect of late de brillation on the outcome. All statistical analyses were performed using the SPSS statistical package, version 24 (IBM). All tests were two-tailed, and P values of less than 0.05 were considered statistically signi cance.

Participant characteristics
During the study period, 2,155 patients with OHCA presumed cardiac etiology who received prehospital de brillation were included. Among them, 165 patients were excluded due to lack of de brillation information. A further 13 patients were excluded due a lack of outcome information (six patients were also missing de brillation information). Finally, a total of 1,983 patients were included in the study and used for analysis.
The median age of the whole population was 61 years, and the majority were male 1,553 (78.3%) ( Table 1). In the prehospital stage, 738 patients underwent ROSC (37.2%). The median age was signi cantly lower in the group of patients with prehospital ROSC than those who were without. The proportion of males were signi cantly higher in the group with ROSC. The proportion of initial shockable rhythm, witnessed cardiac arrest, and bystander CPR were signi cantly higher in the group with ROSC than those without ROSC. The median number of prehospital de brillations was 2 (IQR: 1-4). The median time from arrest to initial de brillation was 10 minutes (IQR: 7-15). The survival rate at hospital discharge was 33.2%, and the proportion of good neurological outcome was 27.7%. Other characteristics were summarized in Table 1. ROSC rates decreased as the number of de brillation attempts at the prehospital stage increased from the rst to the sixth. 16%, 11%, 7%, 5%, 3% and 1%, respectively (Additional Fig). The cumulative ROSC rates and good neurologic outcome from the rst to sixth de brillation were 43%, 68%, 81%, 90%, 95%, 98% and 42%, 66%, 81%, 90%, 95%, 98%, respectively (Fig. 1). A total of 704 (95%) patients showed ROSC with less ve de brillations. Less than ve de brillations were also associated with 95% of good neurologic outcomes. As the time from arrest to rst de brillation affects the patient's outcome, a subgroup analysis was performed on patients who received de brillation within 10 minutes which was the median rst de brillation time. There were 888 patients (44.7%) that had received their rst de brillation within 10 minutes. Similar to the main analysis, it was shown that 95% patients exhibited ROSC with less than ve de brillations (Fig. 2). A subgroup analysis was performed on patients whose rst de brillation time exceeded 10 minutes from cardiac arrest. Similar to the main analysis, the cumulative ROSC rate in patients who underwent de brillation up to ve times was 96% (Fig. 3).

Discussion
This study found that the cumulative prehospital ROSC rates and good neurologic outcome from rst de brillation to the sixth de brillation were 43%, 68%, 81%, 90%, 95%, 98% and 42%, 66%, 81%, 90%, 95%, 98%, respectively. ROSC rates decreased as the number of de brillation attempts at the prehospital stage increased from the rst to the sixth. 16%, 11%, 7%, 5%, 3% and 1%, respectively. More than 95% of prehospital ROSC were achieved within ve de brillation attempts, and more than ve were associated with a sharp decline in the prehospital ROSC rate in OHCA patients. Results from a subgroup of patients that received the initial de brillation in less than 10 minutes, which is the median time from cardiac arrest to rst de brillation, were similar to the main results.
Currently there is no consensus for the number of prehospital de brillation attempts prior to hospital transfer when shockable rhythm is observed. Since the prehospital ROSC rate is reduced after ve or more de brillation attempts, this study may provide a basis for a standardization of the number of de brillation attempts before transfer to hospital. If prehospital ROSC is not achieved, even after four or ve de brillation in the prehospital stage, it is likely that specialized treatment at the hospital, such as extracorporeal membrane oxygenation CPR, could be considered. To the best of our knowledge, the cumulative prehospital ROSC rates based on the number of de brillation attempts using nationwide prospective registry data, especially advanced in recent cardiac arrest resuscitation has not been previously reported.
Early de brillation is vital for the survival of OHCA cases [18,19]. De brillation and chest compression effectiveness decreases rapidly, and survival rates are reduced when the collapse time is > 10 minutes [20]. Several studies have shown that increasing the number of de brillations in OHCA patients is associated with poor prognosis [20,21]. Hagihara et. al. investigated which de brillator type is associated with a better prognosis when the number of de brillations increased, and reported that biphasic de brillators were associated with a better prognosis when compared to monophasic de brillators [20].
There was no analysis of the ideal number of de brillations at the prehospital stage. One study conducted in Japan reported a cutoff point between two and three de brillations, whereby the chance survival after one-month was signi cantly reduced when more than three de brillation were used [21]. In that study, the primary endpoint was one-month survival, and the cumulative outcome rate according to the number of de brillations were not analyzed. Signal detection analysis (determining the largest χ2 value at a certain cut-off point) was used to estimate the ideal number of prehospital de brillations. Since only patients with witnessed cardiac arrests were included, the number of patients used in the nal analysis was limited to 4.2% of the total cohort. There is a difference in nding the ideal number of de brillations that achieve more than 95% of ROSC in our study. Our study demonstrated that increasing numbers of de brillations were signi cantly associated with reduced prehospital ROSC and poorer neurologic outcome. When the time from cardiac arrest to the initial de brillation was too long, de brillation was less effective. Therefore, we performed subgroup analysis using a patient cohort that underwent the rst de brillation within 10 minutes, and resulted in similar outcomes. An alternative to conventional CPR employs extracorporeal membrane oxygenation CPR and it has been reported to be effective for patients who do not t the criteria for conventional CPR [22][23][24]. Patients exhibiting cardiac arrest who are refractory to prehospital de brillation may be considered for further treatment by promptly transferring them to specialized centers. It becomes di cult to decide how many times to perform de brillation in the prehospital stage, but it seems that this study can be a starting point for further research regarding the optimal number of de brillation attempts.
There are several limitations to this study. Our report was driven by prospectively collected registry data, but as it was not created for the purpose of the number or time of de brillations, 165 cases (7%) were excluded from the analysis due to missing de brillation related information. This study did not consider de brillation conducted by bystander. De brillation performed by a bystander can also affect the prehospital ROSC. However, out of all patients, only 40 cases of de brillation were performed by a bystander, so the main result is unlikely to be affected. Another limitation is the type of de brillator in each EMS, whether it is automated external de brillator, monophasic or biphasic, and the energy level will also affect the results, but that information could not be obtained from this registry. However, in Korea, it is known that biphasic de brillators are more commonly used than monophasic. Biphasic de brillators are superior to monophasic in terms of short and long term prognosis of patients with OHCA [20]. We did not address target temperature management or coronary angiography at the hospital stage as adjustment, which may affect patient outcomes. However, since our study's primary endpoint was prehospital ROSC, we do not believe it would have affected the main results. Data regarding cardioactive (epinephrine) and antiarrhythmic agents (amiodarone) were not included in the analysis. However, the frequency of use for was only 57 and 20 cases, respectively, therefore, we do not believe it would have affected the main results.

Conclusion
In our study, an increased number of de brillation attempts at the prehospital stage was signi cantly associated with reduced ROSC. The cumulative ROSC rates from the initial de brillation to the fth de brillation were 43%, 68%, 81%, 90%, and 95%, respectively. For patients with refractory ventricular brillation OHCA, the results of our study have provided a basis for the standardization of the number of de brillations prior to hospital transfer to increase the probability of receiving rescue therapy such as extracorporeal CPR.