Improving Outcomes of Witnessed Out‐of‐Hospital Cardiac Arrest After Implementation of International Liaison Committee on Resuscitation 2010 Consensus: A Nationwide Prospective Observational Population‐Based Study

Between 2007 and 2013, we enrolled 241 990 adult patients with OHCA witnessed by EMS personnel (n=44 706) or bystanders (n=197 284) with initial documented rhythm data from the All‐Japan Utstein Registry (n=841 732; Figure 1). The All‐Japan Utstein Registry is a prospective, nation‐wide, population‐based cohort database of all OHCA patients in Japan. The data set was collected from the standardized style used by the Utstein database.^{3, 4, 5, 6} Among 841 732 OHCA patients, we excluded pediatric patients younger than 18 years (n=12 850), patients without resuscitation attempts (n=13 659), patients missing cardiac origin data (n=3), patients without witnesses to OHCA (n=480 304) or information regarding bystander detail (n=65 098), patients without initial rhythm data (n=27 786), and patients without outcome information (n=42) to identify whether survival from OHCA changed during the 2 sampling periods. Finally, 241 990 patients with witnessed OHCA were enrolled and analyzed in this study (Figure 1). The ethics committee at Osaka University Graduate School of Medicine (Osaka, Japan) approved the study (Approval No. 16064), and the requirement of written informed consent was waived. The study was performed in accord with the Declaration of Helsinki.

Japanese CPR guidelines were developed and released by the Japan CPR Guideline Committee in October 2006 based on the ILCOR 2005 consensus. Thus, there was a 1‐year delay between the publication of the 2005 ILCOR consensus recommendations and the publication of the 2006 Japan CPR guidelines. In October 2010, the Japan CPR guidelines were published simultaneously with the ILCOR 2010 consensus recommendations. Therefore, in this study, we assumed that CPR was performed based on the ILCOR 2005 consensus between 2007 and 2010 (ILCOR 2005 period group) and CPR was performed based on the ILCOR 2010 consensus between 2011 and 2013 (ILCOR 2010 period group), even though there might have been a crossover period between when the new guidelines were introduced and when complete implementation occurred.⁷ Details of the Japanese EMS system have been described elsewhere.^{5, 6} We defined cardiac arrest as the cessation of cardiac mechanical activity confirmed by the absence of signs of circulation.^{3, 4} Diagnoses of cardiac or noncardiac origin were clinically made by physicians in collaboration with EMS personnel by excluding the possibility of noncardiac causes in accord with the Utstein‐style international guidelines.^{3, 4, 5, 6} This means that the diagnosis of cardiac origin was made when there was no evidence of a noncardiac cause, and any other information regarding comorbidities or underlying conditions was not available.^{3, 4, 5, 6}

Japan has an area of ≈378 000 km², and its population was ≈127 million in 2013. There were 770 fire stations with dispatch centers in 2013; EMS is provided by municipal governments. EMS is available 24 hours per day, 365 days per year. When people call emergency services for an ambulance, it will be dispatched from the nearest fire station with 3 emergency providers, with at least 1 highly trained prehospital emergency care provider.

The All‐Japan Utstein registry is a mandatory registry and includes data on all those with EMS‐treated cardiac arrest in Japan.^{5, 6} EMS providers do not cease resuscitation in the field. Resuscitation is attempted for all patients in cardiac arrest unless death is obvious, such as in the case of decapitation, incineration, decomposition, rigor mortis, or the presence of dependent cyanosis.^{5, 6}

We collected the following clinical data from the All‐Japan Utstein Registry: age; sex; enrollment year; origin of arrest (cardiac or noncardiac); type of bystander based on the Utstein Style (family, passerby, friend, and colleague); type of CPR procedure including bystander AED use, which indicates that bystanders delivered shock using AED; first documented cardiac rhythm (shockable or nonshockable); EMS treatments before hospital arrival, such as administration of epinephrine and the use of advanced airway management devices; EMS response time (time from emergency call to first contact with a patient); and hospital arrival time (time from emergency call to patient arrival at a hospital). Both conventional and chest compression–only CPR were considered bystander CPR. The first documented rhythms were classified into 2 categories according to the ILCOR consensus. Those included shockable rhythms defined as pulseless ventricular tachycardia or ventricular fibrillation and nonshockable rhythms, including pulseless electrical activity or asystole, respectively. When bystanders delivered shocks using publicly accessible AEDs, the patient's first documented rhythm was regarded as shockable rhythm.

The primary outcome measure was neurologically favorable 1‐month survival and the secondary outcomes were 1‐month survival and prehospital return of spontaneous circulation (ROSC). Neurological outcome was determined at 1 month after successful resuscitation with the Glasgow–Pittsburg cerebral performance category system: category 1, good cerebral performance; category 2, moderate cerebral disability; category 3, severe cerebral disability; category 4, coma or vegetative state; and category 5, death. Glasgow–Pittsburg cerebral performance scale categories 1 and 2 were defined as neurologically favorable in this study.^{3, 4}

Categorical data are presented as percentage (%), and continuous data are presented as medians (25th–75th percentiles). The chi‐square test or Wilcoxon rank‐sum test was used for comparisons of categorical and continuous data between groups, respectively. We divided our patients into 2 groups: ILCOR 2005 period and ILCOR 2010 period. Then, we evaluated and compared patient backgrounds and outcomes between the 2 groups. For comparisons of the type of bystander (citizen, which consisted of family, passerby, friend, and colleague) and type of CPR procedure (which consisted of conventional, compression‐only, and respiratory‐only), P values from the chi‐square test for contingency table were evaluated. Outcomes associated with the ILCOR 2010 consensus guidelines period were compared to those associated with the ILCOR 2005 consensus guidelines period using a multivariable logistic regression model and presented as adjusted odds ratio (OR) and 95% CI. Adjusted covariates included: age; sex; enrollment year; cardiac versus noncardiac; type of bystander and CPR attempt by bystander, if appropriate; initial rhythm, if appropriate (shockable versus nonshockable); EMS treatments, including epinephrine and advanced airway management; EMS response time; and hospital arrival time; P<0.05 was considered statistically significant. To evaluate the differences among the maturation cycles of the guideline iterations, we performed a multivariable logistic regression analysis including a cross‐product term between the version of ILCOR consensus and the maturation cycle of the guideline. Adjustments were performed similarly to those described for the above logistic regression model. We conducted all statistical analyses by using the R software package (version 3.1.0; R Development Core Team).

This study had some limitations. First, our data did not consider the potential various forms of in‐hospital care, although the use of these treatments could be associated with the improved outcomes. Even though there were no obvious improvements in outcomes of patients who achieved prehospital ROSC in the present study, we speculated that this happened because patients with more‐severe illnesses may have achieved prehospital ROSC in the ILCOR 2010 period as compared with the ILCOR 2005 period. In addition, the number of medical centers providing targeted temperature management, extracorporeal membrane oxygenation for resuscitation, or bundled postcardiac arrest care are unknown. Second, analysis of observational data could not adjust unmeasured confounding factors during the multivariable regression model. Third, there are no data regarding how quickly all EMS providers were retrained within the study period and no data regarding CPR quality assessment during resuscitations, even though there might have been a crossover period between the time when the new guidelines were introduced and when complete implementation occurred.⁷ Fourth, a significant proportion of 1‐month survivors had poor neurological outcomes in the present study compared with other countries²⁵ however, this phenomenon could be explained by differences in EMS systems between Western countries and Japan. In Japan, all OHCA patients who were treated by EMS personnel were transported to a hospital unless death was obvious. In comparison, the proportion of field termination of resuscitation is often at least 50% of all OHCA patients in other countries.²⁶ Fifth, because of the nature of epidemiological studies, validity and integrity are potential limitations of this study. Therefore, our data should be interpreted with consideration of these limitations. However, these potential sources of bias should be minimized by using the uniform data collection style used by the Utstein database.