A Systematic Review of Cost-Effectiveness of Treating Out of Hospital Cardiac Arrest: Implications for Resource-limited Health Systems

Background Out-of-hospital cardiac arrest (OHCA) is a prevalent condition with high mortality and poor outcomes even in settings where extensive emergency care resources are available. Interventions to address OHCA have had limited success, with survival rates below 10% in national samples of high-income countries. In resource-limited settings, where scarcity requires careful priority setting, more data is needed to determine the optimal allocation of resources. Objective To establish the cost-effectiveness of OHCA care and assess the affordability of interventions across income settings. Methods The authors conducted a systematic review of economic evaluations on interventions to address OHCA. Included studies were (1) economic evaluations (beyond a simple costing exercise); and (2) assessed an intervention in the chain of survival for OHCA. Article quality was assessed using the CHEERs checklist and data summarised. Findings were reported by major themes identified by the reviewers. Based upon the results of the cost-effectiveness analyses we then conduct an analysis for the progressive realization of the OHCA chain of survival from the perspective of decision-makers facing resource constraints. Results 468 unique articles were screened, and 46 articles were included for final data abstraction. Studies predominantly used a healthcare sector perspective, modeled for all patients experiencing non-traumatic cardiac OHCA, were based in the US, and presented results in US Dollars. No studies reported results or used model inputs from low-income settings. Progressive realization of the chain of survival could likely begin with investments in TOR protocols, professional prehospital defibrillator use, and CPR training followed by distribution of AEDs in high-density public locations. Finally, other interventions such as indiscriminate defibrillator placement or adrenaline use, would be the lowest priority for early investment. Conclusion Our review found no high-quality evidence on the cost-effectiveness of treating OHCA in low-resource settings. Existing evidence can be utilized to develop a roadmap for the development of a cost-effective approach to OHCA care, however further economic evaluations using context-specific data are crucial to accurately inform prioritization of scarce resources within emergency care in these settings.

approach to OHCA care, however further economic evaluations using context-speci c data are crucial to accurately inform prioritization of scarce resources within emergency care in these settings.

INTRODUCTION
Cardiac arrest is a leading cause of death in high-income countries. 1Given its sudden and often dramatic presentation, this condition draws signi cant public attention.Since the introduction of interventions such as cardiopulmonary resuscitation (CPR) and external de brillation in the mid-20th century, emergency care systems in high-resource settings have devoted substantial effort and funding toward increasing access to treatment for cardiac arrest.A "Chain of Survival" focusing on nontraumatic out-of-hospital cardiac arrest (OHCA) has been conceptualized to include: activation of the emergency care system; bystander/professional CPR; public access/professional de brillation; medications; transport to the facility; and post-arrest care in cases of return of spontaneous circulation (ROSC). 2 Bystander training programs, public awareness campaigns, and public placement of automated external de brillators (AEDs) are examples of public health efforts to improve outcomes.Despite these innovations, outcomes remain poor across the globe.A 2010 systematic review of 67 international studies examining OHCA found a global survival rate of ~ 7%. 3 Globally 30-day survival or survival to hospital discharge remains low: 8.8% in a 2020 systematic review. 4Even in health systems able to devote ample resources towards the chain of survival, the challenges of addressing OHCA have endured.10] Over 80% of the world's population lives in low or middle-income countries (LMICs), but only 20% of global healthcare spending is devoted to this population. 11Few of these healthcare dollars go to emergency care systems, despite evidence that over 50% of the disease burden is attributable to conditions amenable to emergency care. 12In low-resource settings, where scarcity requires careful priority setting, it is unclear whether the investment in interventions to address OHCA represents the best use of resources.In settings where emergency care systems are in early development, it is critical to evaluate the cost-effectiveness of OHCA interventions to better inform resource allocation which maximizes health bene ts.
We conducted a systematic review of the economic evidence surrounding interventions for nontraumatic OHCA in all settings, including speci c interventions in the chain of survival.We then use these results and apply them to our knowledge about low-resource settings to better understand the potential costs and effects of OHCAs in low-resource settings.The results will inform ongoing discussions occurring in countries across the world regarding the allocation of scarce resources within emergency care development.

METHODS
We conducted a systematic review of the literature using the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guidelines. 13The review was accepted and registered with the International prospective register of systematic reviews (PROSPERO-CRD42022310780). Six databases (PubMed, EMBASE, Global Health, Cochrane, Global Index Medicus, and Tuft's Cost-Effectiveness Registry) were searched for articles related to the cost-effectiveness of emergency care interventions to treat non-traumatic OHCA.Search terms were entered in English, without language restriction on the article's language or date restriction.Broad terms that would capture any research around OHCA and its interventions were paired with cost-effectiveness terms.Traumatic cardiac arrest fell outside of the scope of this review.
A sample of PubMed search terms is included in Box 1. Inclusion criteria for an article progressing to data extraction were: (1) An economic evaluation (beyond a simple costing exercise); and (2) an assessment of an intervention in the chain of survival for OHCA.
Box 1: PubMed Search Terms ("Out-of-Hospital Cardiac Arrest"[Mesh] OR "out of hospital cardiac arrest"[tw] OR "out of hospital cardiac arrests"[tw] OR "out of hospital heart arrest"[tw] OR "out of hospital heart arrests"[tw] OR "out of hospital cardiopulmonary arrest"[tw] OR "OHCA"[tw] OR "prehospital cardiopulmonary arrest"[tw] OR "pre-hospital cardiopulmonary arrest"[tw] OR "prehospital cardiac arrest"[tw] OR "pre-hospital cardiac arrest"[tw] OR "prehospital heart arrest"[tw] OR "pre-hospital heart arrest"[tw] OR "automated external de brillator"[tw] OR "automated external de brillators"[tw] OR "bystander cardiopulmonary resuscitation"[tw] OR "bystander CPR"[tw] OR "layperson cardiopulmonary resuscitation"[tw] OR "layperson CPR"[tw]) AND ("Cost-Bene t Analysis"[Mesh] OR "cost bene t"[tw] OR "cost effective"[tw] OR "cost effectiveness" [tw] OR "economic evaluation"[tw] OR "economic evaluations"[tw] OR ("cost"[tw] AND ("bene t"[tw] OR "bene ts"[tw])) OR "cost utility"[tw] OR "cost savings"[mh] OR "cost savings"[tw] OR "cost control"[tw] OR "costs and cost analysis"[mh:noexp] OR "cost analysis"[tw] OR "cost comparison"[tw]) Following PRISMA guidelines, two reviewers independently screened articles for relevance based on title and abstract.Studies that did not use an economic evaluation study design, or were not related to an OHCA care intervention were excluded.Full texts of the remaining articles were then screened again for inclusion eligibility with any con icts requiring resolution by a senior reviewer.The reference lists of all included texts were hand-searched for further potential inclusions.
Included studies were then assessed for quality using the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) guidelines. 14The statement uses 25 checklist items across six main categories ranging from (1) title and abstract; (2) introduction; (3) methods; (4) results; (5) discussion; and (6) other.Scores of this exercise can be found in Appendix 1.Studies were not excluded from our review based on their scores.
For each study that met all inclusion criteria data were abstracted using a prede ned 17-item extraction matrix.To characterize the attributes of included studies, data relevant to economic evaluations were extracted including country of study, intervention and comparator, study design, characteristics of the study population, perspective, time horizon, discount rate, currency and currency year, willingness to pay threshold, and incremental cost-effectiveness ratio.Included studies were rst analysed by theme through discussion and consensus amongst reviewers.We then conducted a decision analysis from the perspective of decision-makers with resource limitations.

Patient and Public Involvement
There was no patient or public involvement in the methods or conduct of this study.

RESULTS
Figure 1 presents our PRISMA ow diagram.478 articles were identi ed in our initial search.After removing ten duplicate studies, 468 studies were assessed on title and abstract.80 full-text articles were then screened by two blinded independent reviewers.During this phase, 38 studies were excluded.The majority of studies (16) were excluded due to not being full of economic evaluations.Six of the 16 were systematic reviews that fell outside of our inclusion criteria; however, in this case the authors searched the full reference lists of all systematic reviews for additional relevant literature.Eight studies were only available as abstracts, and two studies were unavailable in full text.Twelve studies did not describe OHCA interventions and two studies did not assess a health outcome measure.Reference lists were checked for all relevant studies, leading to the identi cation of an additional four studies.46 studies were eligible for data abstraction.

Study characteristics
Descriptive characteristics were identi ed in the included studies and are discussed below.Table 1 includes a summary of the abstracted data from the systematic literature review.

Perspective
The choice of perspective determines which costs and effectiveness outcomes are included in the analysis.Studies in our review used three categories of perspectives including societal, health system/ health care sector, and patients/ payors.While a health sector perspective considers only costs incurred in the provision of medical care, the societal perspective accounts for medical costs as well as lost wages and productivity due to ill health.Best practice recommends the use of societal and healthcare sector perspectives. 15,16Most studies in our review approached their analysis using a healthcare sector perspective (22 articles: 48%), followed by those that used a societal perspective (12 articles; 26%).A single study used the patient-payer perspective, and only two studies conducted their analyses using both a health system and societal perspective. 17,18The remaining studies did not report the perspective used in their analysis (9 articles; 20%).

Intervention and comparator
access external brillators (AEDs) were the most common intervention assessed (18 articles; 39%).Eight studies (17%) compared CPR and bystander training.Studies focusing on pre-hospital interventions compared various attributes of EMS systems capable of resuscitating patients against those without such a capability 19 as well as the value of prehospital critical care. 20Frequently this focused on comparing the deployment of various cadres of support providers including specially trained emergency medical technicians, [21][22][23] police personnel, 24 and re stations 25 , to the current standard of only emergency physicians.The effect of reducing ambulance response time was addressed in two studies. 26,27The use of adrenaline during resuscitation was compared to saline placebo in two studies. 28,29iGel supraglottic airway was assessed compared to tracheal intubation during arrest in two studies. 30,31

Simulated population
The age of the simulated population ranged.Most the included studies did not limit their population to a particular age, modeling for all patients experiencing non-traumatic cardiac OHCA (29 articles; 63%).However, some studies simulated adult-only populations (13 articles; 28%), while only two studies simulated a cohort of children with cardiac conditions. 32,33Furthermore, the characteristics of simulated populations were highly heterogeneous.While some studies speci ed patients by characteristics related to their OHCA, such as patients who were witnessed with a rhythm, 26,34 others speci ed patient populations where there was a lack of witness or EMS present at the time of the arrest. 17Some studies required patients to be unconscious [35][36][37] or that resuscitation had been attempted. 38

Modeling approach
Studies in our review covered a variety of modeling approaches.Most studies used a decision-analytic model (20 articles; 43%).Ten studies used Markov-based models (including decision-analytical Markov models).One study utilized a microsimulation approach. 39However, a fair number of studies did not report on their model approach (9 articles; 20%).

Time horizon
Time horizons used in the models ranged from six months to a lifetime horizon.Fifteen studies (33%) did not report the time horizon used in their model.

Discounting
Discount rates ranged from 1.5-6%.The most common rate chosen was 3% (16 articles; 35%).Three studies where the time horizon of the analysis was 1 year or less did not apply discounting.Fifteen studies (33%) did not report the use of a discount rate.

Incremental Cost-Effectiveness Ratios and Other Cost Analysis Findings
Studies focused primarily on CPR training, public or increased access to de brillation, and prehospital interventions.A single study evaluated the cost-effectiveness of termination of resuscitation (TOR) protocols.A summary of these ndings, with currency standardized to 2022 USD, is included in Table 2.

Quality assessment
0][41] These studies did not report all parameters used in their analysis, nor conducted sensitivity analyses around key variables of uncertainty.However, most studies (25 articles; 54.3%) achieved between above 50-80% of the CHEERs list, and 15 articles (32.6%) could be considered high quality achieving 80% or more of the CHEERS list.

CPR and bystander training
Eight studies on bystander training demonstrated a wide range of cost-effectiveness, from $20,534 to $298,727 USD (2022)/QALY gained.Training laypersons was explored through two studies. 43,45Two compared the cost-effectiveness of CPR combined with AEDs against basic CPR 37,47 while one investigated the use of mechanical CPR versus manual chest compressions. 46The results were dependent on factors such as equipment costs (the use of mannequins and AED trainers) and training venues (small size vs mass training).The ICER was highest in the case of trainees living with high-risk individuals over the age of 75; $3,674,607/QALY. 45

Public Access De brillation
Eighteen studies were identi ed that presented data related to public access de brillation.Often public access AEDs were compared against the current standard which was no AEDs, however, in one case, the use of AEDs by laypersons was compared to the use of AEDs by EMS personnel. 64Three studies targeted the placement of AEDs in residences of high-risk populations including adults over 60, longterm care facilities, and children with heart conditions. 32,50,52Two papers compared igel supraglottic airway (SGA) to tracheal intubation, however, one did not nd any evidence of a difference in cost-effectiveness, and another found that igel SGA was less effective and more costly than tracheal intubation. 31 single study was identi ed in our review on the economic value of TOR protocols. 63When comparing BLS with TOR to a situation with no TOR, it was determined that the no TOR rule scenario was not costeffective.Among three scenarios (BLS with TOR, ALS with TOR and no TOR) the BLS with TOR protocols was most cost-effective with an ICER of $23,851/QALY.Given the function of TOR as an administrative tool to reduce resource waste, it can be hypothesized that they would be cost-effective or cost-saving in most settings. 65 use the results identi ed in our review to develop a resource prioritization league  69 Given the high mortality and morbidity, even when maximal resources are leveraged, questions have been raised regarding the ethics and appropriateness of prioritizing OHCA, particularly in low-resource settings. 12The opportunity cost of delaying the implementation of competing interventions that are less costly but more e cacious becomes a concern under tight scal constraints.
Progressive realization of the chain of survival could likely begin with investments in TOR protocols, professional prehospital de brillator use, and CPR training, followed by distribution of AEDs in highdensity public locations.Finally, other interventions such as indiscriminate de brillator placement, or adrenaline use would be the lowest priority for early investment.
Unfortunately, our review yielded only two studies on the OHCA chain of survival in an middle-income setting, both in upper-middle-income settings (Colombia and Mexico).No evidence was found from the most resource-limited settings, hindering the transferability of the results, and highlighting the acute need for more research to be conducted in low-resource contexts.Caution should be taken in extrapolating results from high-income to low-income settings for a variety of reasons including differences in health systems, the values of parameters, and cost-effectiveness thresholds among these settings.Relying on results representative of high-income country (HIC) data also means that some costs, particularly human resources, are in ated compared to a low or middle income (LMIC) setting.However, given the more established emergency care infrastructure in HICs, interventions may perform better and achieve greater outcomes than they would in LMIC emergency care naïve settings as well.
Given the lack of transferability of these ndings to other settings and the context-speci c challenges, more primary research is needed for OHCA in low-resourced settings.
There are several other limitations important to consider for decision-makers wishing to utilize this evidence.For one, publication bias may favour studies with positive results, therefore our analyses may overestimate the bene t of the themes identi ed in our review.Furthermore, while many studies attempted to isolate the effect of certain elements of the chain of survival, all studies occurred in the presence of pre-existing capacities that may have in uenced outcomes or led to an underestimation of costs.For example, studies on the cost-effectiveness of bystander CPR in a high-income country make no attempts to control for the existing ambulance and hospital resources that exist in the status-quo base-case comparator.
Care must also be taken in comparing cost-effectiveness ratios across a wide variety of models or intervention intensities.For example, the cost-effectiveness of public access de brillation varied signi cantly based on how targeted the placement of AEDs was, and the cost-effectiveness of prehospital interventions varied widely based on urban versus rural population distribution.Furthermore, the interventions identi ed in our review are highly heterogeneous and presented incomplete methods in terms of time horizon, modeling approach, and perspective used.Finally, we do not recommend that economic considerations serve as the sole factor behind priority setting, resource allocation, or policy development.

CONCLUSION
OHCA is a prevalent condition with high mortality and poor outcomes even in settings where extensive resources are devoted to this condition.In settings of resource scarcity where allocation decisions inevitably lead to healthcare rationing, economic analyses can aid prioritization.Our systematic review found 46 studies evaluating the cost-effectiveness of treating non-traumatic OHCA.Our review found no high-quality evidence on the cost-effectiveness of treating OHCA in resource-limited settings.Existing evidence can be utilised to develop a roadmap for the development of a cost-effective approach to OHCA care.However further economic evaluations should be completed using context-speci c data to help accurately inform prioritization and resource allocation decisions in these settings.

Declarations
Ethics approval consent to participate : Not applicable Consent for publication: Not applicable Availability of data and materials: All data generated or analysed during this study are included in this published article and its supplementary information les Competing interests: The authors declare that they have no competing interests Funding: None Authors' contributions: NR, LW, and HG conceptualized the study.NR, KW, OO, SH participated in the systematic review, data collection, and analysis.EH and JG prepared data visualizations.All authors contributed equally to study design, data interpretation and writing.
. Incremental cost-effectiveness results of included studies by theme] 32,27,based on the likely cost-effectiveness of a progressive realization of interventions in the chain of survival.Figure 1 illustrates the league table illustrating which interventions could be considered appropriate for prioritized funding.The guide comprises a list of interventions in ascending order (from low to high) based on ICER.This represents a prioritized list of interventions, from high priority (low cost per QALY) to low priority (high cost per QALY) for the aim of generating as many QALYs as possible.Using the reported ICERs identi ed in our review, we develop prioritization recommendations to rank the most valuable elements in the chain of survival based on the generation of QALYs per investment.This is particularly useful for decision-makers working under resource constraints.OHCA interventions ranked from highest to lowest priority include TOR protocols, AED for EMS, improvements to prehospital systems, CPR bystander training, and lastly public access AED placement.Three studies comparing public access versus targeted in-home AED placement under the same context and setting determined targeted placement of AEDs with high-risk patients to be more cost-effective than indiscriminate public access deployment.32,49,56Healtheconomicsevaluations,typicallyrepresented in the form of cost-effectiveness studies, provide a quantitative tool to help inform resource allocation.While economics should never be the sole driver behind policy or resource decisions, comparing two or more interventions on the basis of costs and effects plays a critical role in priority-setting activities.Should decision-makers be faced with a choice of which of the interventions to adopt rst, given limited resources, evidence suggests certain parts of the chain of survival may yield higher health returns for nite resources.Increased public access to AEDs is likely cost-effective, but only when placed in high-density public locations.De brillator use by professionals in the prehospital system is likely cost-effective as well.The addition of bystander training in CPR and automated external de brillators (AEDs) is likely cost-effective for fewer contexts.If we were to apply the ndings of this review to lower resource settings, using the World Health Organization's de nition of cost-effectiveness for health interventions (less than three times the annual gross domestic product (GDP) per capita of a country for each disability-adjusted life year (DALY) averted or quality-adjusted life-year (QALY) saved),26,27for the purpose of health system planning, a majority of the interventions identi ed in our review would not be recommended in low and lower-middle income countries, with many interventions only being recommended in the high-income setting.It is likely that resources would yield a greater impact if directed towards alternate health interventions, in situations where mutually exclusive allocation decisions must be made in the short term.Cost-effectiveness evidence does exist for numerous emergency care interventions in low-resource settings.66Forexample, training lay rst responders in basic trauma care in urban Uganda costs $25-$75 USD/life year saved, 67 prehospital electrocardiograms for patients with acute chest pain in India costs $12.65 USD per QALY gained 68 and basic paediatric emergency care training and triage costs $148 USD per death averted.
DISCUSSIONThis systematic review described and summarized the published evidence related to the costeffectiveness of non-traumatic OHCA interventions.We highlight studies that identify cost-effective interventions across four areas of the chain of survival including (1) the use of automated external de brillators, (2) CPR and bystander training, (3) prehospital system interventions, and (4) termination of resuscitation guidelines.Studies predominantly used a healthcare sector perspective, modeled for all patients experiencing non-traumatic cardiac OHCA, were based in the US, and presented results in US Dollars.No studies reported results or used model inputs from low-income settings.