Impact of Cardiac Arrest Centers on the Survival of Patients With Nontraumatic Out‐of‐Hospital Cardiac Arrest: A Systematic Review and Meta‐Analysis

Background The role of cardiac arrest centers (CACs) in out‐of‐hospital cardiac arrest care systems is continuously evolving. Interpretation of existing literature is limited by heterogeneity in CAC characteristics and types of patients transported to CACs. This study assesses the impact of CACs on survival in out‐of‐hospital cardiac arrest according to varying definitions of CAC and prespecified subgroups. Methods and Results Electronic databases were searched from inception to March 9, 2021 for relevant studies. Centers were considered CACs if self‐declared by study authors and capable of relevant interventions. Main outcomes were survival and neurologically favorable survival at hospital discharge or 30 days. Meta‐analyses were performed for adjusted odds ratio (aOR) and crude odds ratios. Thirty‐six studies were analyzed. Survival with favorable neurological outcome significantly improved with treatment at CACs (aOR, 1.85 [95% CI, 1.52–2.26]), even when including high‐volume centers (aOR, 1.50 [95% CI, 1.18–1.91]) or including improved‐care centers (aOR, 2.13 [95% CI, 1.75–2.59]) as CACs. Survival significantly increased with treatment at CACs (aOR, 1.92 [95% CI, 1.59–2.32]), even when including high‐volume centers (aOR, 1.74 [95% CI, 1.38–2.18]) or when including improved‐care centers (aOR, 1.97 [95% CI, 1.71–2.26]) as CACs. The treatment effect was more pronounced among patients with shockable rhythm (P=0.006) and without prehospital return of spontaneous circulation (P=0.005). Conclusions were robust to sensitivity analyses, with no publication bias detected. Conclusions Care at CACs was associated with improved survival and neurological outcomes for patients with nontraumatic out‐of‐hospital cardiac arrest regardless of varying CAC definitions. Patients with shockable rhythms and those without prehospital return of spontaneous circulation benefited more from CACs. Evidence for bypassing hospitals or interhospital transfer remains inconclusive.


Yeo et al
Cardiac Arrest Centres in Out-of-Hospital Arrest care has been suggested to be the fifth link in the chain of survival concepts, and a component of an integrated emergency care network comprising community first responders, emergency medical services (EMS), and hospitals aiming to provide quality care to patients with OHCA. Despite advances in therapeutics such as targeted temperature management (TTM), mechanical circulatory support, and neuroprognostication, urgent questions remain pertaining to how best to organize hospitals and emergency care systems to improve access to quality care and clinical outcomes. 8 The recent 2020 guidelines from the International Liaison Committee on Resuscitation recommended with low certainty 9-11 that patients with OHCA should be transported to cardiac arrest centers (CACs). CACs are specialized tertiary institutions, conceptually similar to level 1 trauma centers, and are often high-volume or regionalized centers treating patients with OHCA with the capability to organize postresuscitation care, including 24/7 access to a cardiac catheterization laboratory for coronary angiography and percutaneous coronary intervention (PCI), TTM, extracorporeal membrane oxygenation, and neuroprognostication among other interventions. 12-14 However, although there has been evidence for the effectiveness of each individual intervention in variable settings, 15-19 evidence for the benefit of CACs in treating patients with OHCA remain inconclusive. This is in part because CACs, which provide a complex bundle of interventions, have been poorly defined, 10,20 and similar institutions described in published literature may range from exhibiting only a few to many of the defining traits of a CAC. This brings about difficulties in statistical analysis and interpretation.
Furthermore, it is unknown which subpopulations benefit more from CACs, defined according to prehospital Utstein variables 21,22 such as the receipt of bystander cardiopulmonary resuscitation, initial shockable rhythm, or prehospital return of spontaneous circulation (ROSC). A recent cohort study by Chien et al 23 suggested that the presence of a shockable rhythm modified the benefits of CAC. Kajino et al 24 also showed significant benefit in patients without, but not for patients with, prehospital ROSC. This knowledge gap is especially pertinent because knowing which patients are likely to benefit from an expensive intervention can guide prioritization of scarce health care resources. This understanding also aids the rational ambulance diversion strategy to bring the right patients to the right destinations. 8,25 Consequently, this systematic review aimed not only to provide urgently needed evidence for or against treating patients with OHCA at CACs, but also to analyze the impact within predefined subgroups.

METHODS
This systematic review and meta-analysis adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (Table S3). 26,27 The study protocol had been published in the PROSPERO (International Prospective Register of Systematic Reviews; CRD42021260468). The data that support the findings of this study are available from the corresponding author upon reasonable request.

CLINICAL PERSPECTIVE
What Is New?
• There is uncertainty over the role of cardiac arrest centers (CACs) in the care of out-of-hospital cardiac arrest (OHCA), and the 2020 International Liaison Committee on Resuscitation guidelines previously recommended with low certainty that patients with OHCA should be transported to a CAC, partly based on a systematic review on the topic. • Treatment of nontraumatic patients with OHCA at CACs was associated with significantly improved survival and neurological outcomes, and these findings persisted even when using varying definitions of CAC (eg, high-volume centers). • The treatment effect was more pronounced among patients with OHCA with shockable rhythm and those without prehospital return of spontaneous circulation.
What Are the Clinical Implications?
• The current updated systematic review and meta-analysis provided an upgraded level of evidence (Grading of Recommendations, Assessment, Development, and Evaluation level of evidence: moderate) in support of transport of nontraumatic patients with OHCA to CACs, and patients who would likely benefit most are those with shockable rhythms and those without prehospital return of spontaneous circulation. • Regionalized care for patients with OHCA has the potential to improve outcomes, but transport policies that involve bypassing the nearest hospital for CACs or for interhospital transfer from non-CACs to CACs need further studies.

Search Strategy
A systematic literature search was performed in MEDLINE, Embase, and Cochrane Central Register of Controlled Trials databases from inception up to March 9, 2021. The search strategy was developed in consultation with a medical information specialist. Keywords and Medical Subject Headings terms such as "cardiac arrest center," "hospital volume," "postresuscitation care," "fifth link," "out-of-hospital cardiac arrest," and other synonyms were applied in the search strategy to identify relevant articles. Seventy-one references from the latest systematic review and meta-analysis on this topic, 10 including the review itself, were hand searched to identify additional relevant studies. The investigative team, which included several resuscitation scientists, were asked to ascertain if they were aware of additional relevant studies. This process did not surface any study that was not already captured in the search strategy. Articles were viewed through Endnote X9 28 (Clarivate, Philadelphia, PA) for an article sieve. The search was repeated on June 7, 2021 yielding no additional eligible articles. The detailed search strategy is available in Data S1.

Definition of Cardiac Arrest Centers
There was a lack of consensus over the definition of a CAC in the literature. 7,10,12,20 For example, the Association for Acute CardioVascular Care of the European Society of Cardiology described cardiac arrest centers 12 as specialized institutions offering all recommended treatment options for patients with OHCA, including access to a coronary angiography laboratory with 24/7 PCI capability, TTM, extracorporeal membrane oxygenation, mechanical ventilation, and neurological prognostication. On the other hand, the German Resuscitation Council accreditation process required CACs to have standard operating procedures for communication with EMS and quality of care assessments in addition to 24/7 PCI, TTM, and intensive care capabilities. 29 Using a strict definition of CAC, only institutions with the capability for 2 or more of the above interventions and explicitly referred to by study authors as CACs or synonymous terms, such as critical care medical center, tertiary heart center, cardiac receiving center, and regional center, were accepted. PCI-capable hospital alone was not accepted as a term synonymous with CAC. Having the capability for only 1 of the above interventions was also considered insufficient, because a single intervention cannot constitute an intervention bundle. To account for differences in defining CACs, sensitivity analyses were conducted using less strict definitions, accepting terms like high-volume centers and centers with improved postresuscitation care including before-and-after study designs.

Statistical Analysis
Data on general article information (author, year, country), baseline demographics (age, sex, witnessed arrest, initial shockable rhythm, prehospital ROSC), and outcomes of interest (survival to 30 days or hospital discharge with good neurological outcome, survival to 30 days or hospital discharge) were abstracted by 3 authors (J.W.Y., Z.H.C.N., A.X.C.G.). The data abstraction process was blinded among the authors, who used a predetermined data collection form. Disputes were resolved through consensus from the senior author (A.F.W.H.). Adjusted odds ratio (aOR) and crude odds ratio (OR) for binary outcomes were abstracted from each article. Where incremental or hierarchical statistical models were presented, the OR adjusted for the maximum number of covariates was extracted.
Where multiple statistical approaches were presented (eg, multivariable modeling and propensity-score matching) in the same study, we considered the approach used in the primary analysis. When unavailable, OR and 95% CI were calculated for articles reporting summary data using 2×2 contingency tables. Conventional pairwise meta-analyses were performed. Given the high known concordance 30 between the outcomes of survival to 30 days and survival to hospital discharge, the decision was made in consensus with all study authors to pool both outcomes, which was deemed sufficient to demonstrate improvement in short-term OHCA outcomes, if any, consistent with the Core Outcome Set for Cardiac Arrest. 31 The aORs were preferentially analyzed over ORs, because the estimates represent less bias caused by confounding. A DerSimonian-Laird random-effects model with inverse variance weights was applied regardless of heterogeneity because of expected between-study variations in population and interventions. Sensitivity analyses were performed for wider definitions of a CAC, including CACs defined as strictly explicit CACs and high-volume centers, or CACs and centers with improved postresuscitation care. Further sensitivity analyses applying fixed-effects models to the above were also performed. Heterogeneity was assessed using the I 2 statistic with 25%, 50%, and 75% thresholds for low, moderate, and high levels of heterogeneity, respectively. To account for heterogeneity, subgroup analyses were performed to compare studies measuring outcomes by hospital discharge versus 30 days, and also among all included studies for predefined, clinically important Utstein variables: initial shockable rhythm and presence of prehospital ROSC whenever possible. All analyses were performed using Review Manager (RevMan 5.4) software package 32 by the Cochrane Collaboration. Two-tailed statistical significance was set at P<0.05. Publication bias was assessed through visually inspecting funnel plots when 10 or more studies reported an outcome. The quality of observational studies was evaluated on the Newcastle-Ottawa scale, 33

Literature Retrieval and Summary of Included Articles
The database search yielded 4544 articles. There were 1093 duplicate articles removed, and 3358 articles were excluded on the basis of their titles and abstracts. A further 54 articles were excluded upon full-text review. The κ value measuring interrater reliability was 0.75 when reviewing the title and abstracts and 0.9 for full-text review. Finally, 36 studies 20,23-25,37-68 qualified for analysis. The study selection process and reasons for excluding the 54 studies are illustrated in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses-P 2020 flow diagram ( Figure 1).
A total of 147 943 patients were included in the 36 studies. Two studies 55,67 were conducted in Australia, 1 in Canada 66 , 1 in the Czech Republic, 60 2 in Denmark, 47,53 1 in France, 56 59 and 9 § in the United States. Three articles reported data from the CARES (Cardiac Arrest Registry to Enhance Survival) registry, 2 articles from the CAVAS database (Cardiovascular Disease Surveillance), 2 articles from the NHIRD database (National Health Insurance Research), and 3 articles from the UOP (Utstein Osaka Project). Fifteen studies were prospective cohorts, 20 were retrospective cohorts, and 1 was a pilot study for a randomized clinical trial.
The characteristics and quality assessment of included studies are presented in Table S1. The figures for unadjusted analyses are presented in Figure S1. The summary of meta-analysis results is presented in Table. Survival to 30 Days or Hospital Discharge With Good Neurological Outcome

Unadjusted Analyses
Seven studies 23

Subgroup Analysis
Subgroup analysis comparing patients with initial shockable or nonshockable rhythm ( Figure 4)

Subgroup Analysis
Subgroup analysis comparing patients with or without prehospital ROSC (Figure 7)

Direct to CAC Versus Transfer to CAC
Only 2 studies 44,49 reported outcomes for patients directly transported to a CAC versus transferred to a CAC from another hospital. The studies were too heterogeneous to pool, but both reported no significant differences in survival or neurological outcomes between patients directly transported or transferred to a CAC.

DISCUSSION
The optimal CAC configuration and the benefit of CACs on the survival outcomes of patients with OHCA remain uncertain, 69 especially for predefined patient subgroups. Only low-certainty evidence for improved survival in CACs has been demonstrated by previous meta-analyses (Table S2), 7,10,70 which were limited by inconsistencies in CAC definitions and the reliance on before-and-after study designs vulnerable to inherent biases. This is, to our knowledge, the most up-to-date systematic review and meta-analysis on the topic, with 7 new studies since the last review by Yeung et al, and the first to demonstrate clearly improved survival among patients with OHCA treated in CACs compared with non-CACs. The results showed (1) significantly improved survival to 30 days or discharge with good neurological outcome and (2) improved survival to 30 days or discharge for patients with OHCA who received care at a CAC (main analysis), regardless of how strictly CACs were defined (sensitivity analyses). Additionally, subgroup analysis suggested that the treatment effect of CACs may be significantly better for patients with shockable rhythm and without prehospital ROSC. On the whole, only 6 studies using beforeand-after designs were included in this review. These were excluded from the main analysis but included for the sensitivity and subgroup analyses. Taken together, these findings hold implications for the organization of emergency care systems and ambulance diversion strategies for patients with OHCA. High case volume and aggressive postresuscitation care have been shown to improve outcomes for OHCA, both of which are key features of CACs. 12,61,62 This analysis demonstrates improved survival and survival with good neurological outcomes both at discharge and 30 days for patients with OHCA treated at CACs, in contrast to a previous meta-analysis. 10, 70 The quality of included articles using the strict definition of CAC was assessed to be high (≥7), and all were large cohort studies that comprehensively controlled for confounding and did not rely on study designs with historical controls. There was an observed decrease in benefit  when including high-volume centers as CACs as compared with including improved postresuscitation care centers as CACs, possibly suggesting the relatively higher contribution of postresuscitation interventions to the treatment effect, perhaps because cardiac causes of OHCA predominate in this study. 6,65 However, the consistency of significant benefit across all definitions of CAC indicates that patients should be transported to CACs or even hospitals exhibiting the variable features associated with CACs to improve outcomes.   The 2020 International Liaison Committee on Resuscitation statement also noted that evidence for CACs among subgroups of patients remain inconclusive. 69 This analysis contributes to this active debate by demonstrating that survival with good neurological outcomes was more pronounced among patients with shockable rhythm and that survival was more pronounced among patients without prehospital ROSC when comparing transport to CACs and non-CACs. Patients with shockable rhythm have also been associated with OHCA of cardiac causes and may benefit the most from early access to PCI 71,72 and intensive cardiac care. 65 Increased benefit in patients without ROSC also partially supports the view favoring quicker transport of patients with refractory OHCA to a hospital 73-76 instead of prolonging onscene resuscitation, 77 allowing patients to access advanced critical care and extracorporeal membrane oxygenation. These findings should be interpreted with caution, because the studies included for subgroup analysis were vulnerable to bias but offer preliminary evidence that EMS may consider prioritizing patients with shockable rhythm or without prehospital ROSC for transport to CACs. It should also be considered that patients with nonshockable rhythms inherently have poorer survival and neurological outcomes compared with those with shockable rhythms, which may have contributed to findings of poorer survival and neurological outcomes among such patients regardless of the effect of CACs in relation to non-CACs. In addition, because patients with nonshockable rhythms still significantly benefited from a CAC albeit to a lesser extent, this analysis does not support depriving these patients from CAC care, but rather, warrants further examination of the associated incremental cost-effectiveness.
Although transport to a CAC improves outcomes, it remains unclear if EMS should bypass the nearest emergency departments in favor of CACs. 57,78 It has been suggested that the increase in transport time caused by bypassing the nearest hospital does not substantially affect outcomes after transport to CACs. 23,42,57,76,79 Other options include initial transport to a non-CAC with eventual interhospital transfer to a CAC, which seemed to have similar outcomes in this review, 44,49 but more definitive evidence is required to confirm this finding, in the form of an interventional trial comparing ambulance diversion strategies.

Strengths and Limitations
This is the largest systematic review and meta-analysis of evidence for the benefits of CACs conducted to date, involving OHCA registries and databases from However, differences across geographical regions may have led to the high statistical heterogeneity encountered in various analyses. Outcomes at both 30 days and at discharge were pooled, also contributing to heterogeneity, because time points for discharge may differ based on health system. However, we found that our results remained consistent both on pooling survival at 30 days with survival at discharge, and also when examining each of these separately. Another limitation is that the included studies used largely similar but not identical covariates for the adjustment of ORs, which may lead to residual confounding. As a whole, the interpretation of our results should also consider that specific levels of care at non-CAC hospitals were inconsistently defined. Subgroup analyses should be interpreted carefully given existing selection bias by EMS and smaller sample sizes. The conclusion that patients with shockable rhythms do better when transported to CACs may have been driven by a higher proportion of STsegment-elevation myocardial infarction within this group, hence accounting for better outcomes thanks to the presence of cardiac catheterization laboratories in CACs. Furthermore, observational studies are inherently susceptible to selection and observation biases. High-quality randomized clinical trials are therefore urgently needed to confirm present findings. Evidence for direct transport or transfer to a CAC was inconclusively assessed. Non-English language articles were also excluded.

CONCLUSIONS
CACs improved survival and neurological outcomes at discharge or 30 days among patients with OHCA, regardless of how CACs were defined. There was preliminary evidence for EMS to consider transport to CACs, especially for patients with shockable rhythm or patients without prehospital ROSC. High-quality data are needed to confirm these findings and conclusively assess whether patients should bypass the nearest hospital to be transported to a CAC versus transferred to a CAC from the nearest hospital.

Disclosures
Dr Ong reports funding from the Zoll Medical Corporation for a study involving mechanical cardiopulmonary resuscitation devices and an advisory relationship with Global Healthcare SG, a commercial entity that manufactures cooling devices. The remaining authors have no disclosures to report.

Supplementary Material
Data S1 Tables S1-S4 Figures S1-S2    Studies implemented a structured care pathway, defined as an organized treatment protocol which was determined a priori, implemented postcardiac arrest and during the acute hospitalization, and had more than one intervention (e.g., PCI and TTM).
Our findings support a highly organized approach to postcardiac arrest care, in which a cluster of evidence-based interventions are delivered by a specialized interdisciplinary team. Given the overall low certainty of evidence, however, definitive recommendations will depend on confirmation in additional high-quality studies. Additionally, results presented here provide the rationale for future studies which will test optimal combinations and timing of interventions, and which will integrate a structured approach to neurologic prognostication with the goal of further ameliorating care and outcomes in this population. There is also a need for new research which accounts for heterogeneity inherent in this population and validates personalized approaches based on biological subtypes. CAC: Cardiac arrest centers; PCI: Percutaneous coronary intervention; TTM: Targeted temperature management Selection process 8 Specify the methods used to decide whether a study met the inclusion criteria of the review, including how many reviewers screened each record and each report retrieved, whether they worked independently, and if applicable, details of automation tools used in the process.

Page 11
Data collection process 9 Specify the methods used to collect data from reports, including how many reviewers collected data from each report, whether they worked independently, any processes for obtaining or confirming data from study investigators, and if applicable, details of automation tools used in the process.
Page 12-14 Data items 10a List and define all outcomes for which data were sought. Specify whether all results that were compatible with each outcome domain in each study were sought (e.g. for all measures, time points, analyses), and if not, the methods used to decide which results to collect.
Page 12 10b List and define all other variables for which data were sought (e.g. participant and intervention characteristics, funding sources). Describe any assumptions made about any missing or unclear information.
Page 12; Table 1 Study risk of bias assessment 11 Specify the methods used to assess risk of bias in the included studies, including details of the tool(s) used, how many reviewers assessed each study and whether they worked independently, and if applicable, details of automation tools used in the process.
Page 13-14 Effect measures 12 Specify for each outcome the effect measure(s) (e.g. risk ratio, mean difference) used in the synthesis or presentation of results. Page 12-13 Synthesis methods 13a Describe the processes used to decide which studies were eligible for each synthesis (e.g. tabulating the study intervention characteristics and comparing against the planned groups for each synthesis (item #5)).
Page 11-12 13b Describe any methods required to prepare the data for presentation or synthesis, such as handling of missing summary statistics, or data Page 12-13

Item # Checklist item
Location where item is reported conversions.
13c Describe any methods used to tabulate or visually display results of individual studies and syntheses.
Page 13 13d Describe any methods used to synthesize results and provide a rationale for the choice(s). If meta-analysis was performed, describe the model(s), method(s) to identify the presence and extent of statistical heterogeneity, and software package(s) used.
Page 12-13 13e Describe any methods used to explore possible causes of heterogeneity among study results (e.g. subgroup analysis, meta-regression). Page 13 13f Describe any sensitivity analyses conducted to assess robustness of the synthesized results. Page 13 Reporting bias assessment 14 Describe any methods used to assess risk of bias due to missing results in a synthesis (arising from reporting biases). Page 13 Certainty assessment 15 Describe any methods used to assess certainty (or confidence) in the body of evidence for an outcome. Page 13-14

RESULTS
Study selection 16a Describe the results of the search and selection process, from the number of records identified in the search to the number of studies included in the review, ideally using a flow diagram. Figure 1 16b Cite studies that might appear to meet the inclusion criteria, but which were excluded, and explain why they were excluded. Figure 1 Study characteristics 17 Cite each included study and present its characteristics. Supp Material 2 Risk of bias in studies 18 Present assessments of risk of bias for each included study. Supp Material 2

Results of individual studies
19 For all outcomes, present, for each study: (a) summary statistics for each group (where appropriate) and (b) an effect estimate and its precision (e.g. confidence/credible interval), ideally using structured tables or plots. 20b Present results of all statistical syntheses conducted. If meta-analysis was done, present for each the summary estimate and its precision (e.g. confidence/credible interval) and measures of statistical heterogeneity. If comparing groups, describe the direction of the effect.