Study Design and Setting
This study was a retrospective analysis of the JAAM-OHCA registry [8]. The study period was from June 2014 to December 2017. The complete study methodology has been previously described [8,9]. In brief, pre-hospital data collection was conducted by emergency medical service (EMS) personnel according to the Utstein-style template [10,11], whereas the physicians of each participating institution collected in-hospital data, including information on the etiology of arrests, treatments, and outcomes. The registry is ongoing without setting the end date of the registry period. The registry was approved by the ethics committee of each participating hospital.
Study Patients
This study included adult patients aged ≥18 years with OHCA for whom resuscitation was attempted by bystanders or EMS personnel who were then transported to participating institutions without achieving a return of spontaneous circulation (ROSC) upon hospital arrival. The exclusion criteria were patients (1) for whom resuscitation was not attempted, (2) whose pre-hospital data were not available, and (2) whose pre-ROSC pCO2 value was not available.
EMS System in Japan
A detailed description of the EMS system in Japan has been provided previously. The EMS system in Japan has been provided elsewhere [12,13]. Emergency services are provided 24 h a day. After receiving an emergency call, an ambulance is dispatched from the nearest center. The most highly trained emergency care providers are called emergency life-saving technicians (ELSTs) and permit the insertion of an intravenous line and supraglottic airway device for patients with OHCA. In addition, specially trained ELSTs were permitted to perform tracheal intubation and administer intravenous adrenaline. Generally, a crew of three emergency providers, including at least one ELST, is in each ambulance. CPR is performed based on the Japanese CPR guidelines, which are similar to other international guidelines, especially regarding the ventilation strategy: 30:2 compression-ventilation ratio if there is no advanced airway and 10 breaths/min after advanced airway placement [14]. In Japan, termination of resuscitation in pre-hospital settings by EMS personnel is prohibited [14,15]. Therefore, except for cases of decapitation, incineration, decomposition, rigor mortis, or dependent cyanosis, EMS personnel transport almost all patients with OHCA to the hospitals.
Data Collection and Primary Exposure
We collected the following data from the JAAM-OHCA registry: age, sex, causes of arrest, witness status, bystander CPR, a first documented rhythm at the scene or after hospital arrival, pre-hospital epinephrine administration, pre-hospital airway management, blood gas analysis, actual treatments for patients with OHCA (e.g., targeted temperature management, coronary angiography, and percutaneous coronary intervention), and outcome data. The etiologies of cardiac arrest were divided into three categories: cardiac (or presumed cardiac) and respiratory (respiratory disease, hanging, drowning, or asphyxia), or others. Respiratory etiology was kept as a separate category because patients with these etiologies of cardiac arrest might have a basic high pCO2 level; therefore, they should be analyzed separately. The presumed cardiac cause category was determined by exclusion (i.e., the diagnosis was made when there was no evidence of a noncardiac cause) [10,11].
The primary exposure in this study was the intra-arrest pCO2 value obtained from a blood gas analysis performed upon hospital arrival. The patients were categorized into four quartiles based on their intra-arrest pCO2 levels: quartiles 1 (<66.0 mmHg), 2 (66.1–87.2 mmHg), 3 (87.3–113.5 mmHg), and 4 (≥113.6 mmHg).
Outcome Measurements
The primary outcome of this study was 1-month survival with a neurologically favorable outcome. The secondary outcome was 1-month survival. The neurological status of the survivors was evaluated by medical staff at each institution 1 month after the event. Neurologically favorable outcomes were defined as cerebral performance category (CPC) 1 or 2 [10,11,16].
Statistical Analyses
Patient characteristics and outcomes were compared among the four groups using the Kruskal–Wallis rank test for continuous variables and the chi-square test for categorical variables. To evaluate the association between intra-arrest pCO2 levels and favorable neurological outcomes, we applied univariable and multivariable logistic regression analyses with odds ratios as the effect variables. We selected the following potential confounders: age (continuous), sex (male or female), etiology of cardiac arrest (cardiac, respiratory, or other), bystander witness (yes or no), bystander CPR (yes or no), first documented rhythm (shockable, non-shockable, or other), pre-hospital advanced airway method (bag-mask ventilation, supraglottic airway, or tracheal intubation), pH level, the interval between the call, and time of blood gas sampling [17-20]. As stratified analyses using the method of ventilation and etiology of arrests, we visually described the nonlinear relationship between the intra-arrest pCO2 level and the estimated probability of favorable neurological outcome using a restricted cubic spline with the univariable logistic regression model. In addition, to minimize the impact of “no ventilation time,” we focused on witnessed arrests and performed univariable and multivariable logistic regression analyses using the same covariates as the main ones. All P-values were two-sided, and values <0.05 were considered statistically significant. All statistical analyses were performed using Statistical Product and Service Solutions (version 25 J; IBM Corp. Armonk, NY, USA) and R software (R Foundation for Statistical Computing, version 3.4.3).