Out of hospital cardiac arrest in Western Sydney - an analysis of outcomes and estimation of future eCPR eligibility


 Background

Refractory out of hospital cardiac arrest (OHCA) is associated with extremely poor outcomes. However, in selected patients extracorporeal cardiopulmonary resuscitation (eCPR) may be an effective rescue therapy, allowing time treat reversible causes. The primary goal was to estimate the potential future caseload of eCPR at historically 'low-volume' extracorporeal membrane oxygenation (ECMO) centres.
Methods

A 3-year observational study of OHCA presenting to the Emergency Department (ED of an urban referral centre without historical protocolised use of eCPR. Demographics and standard Utstein outcomes are reported. Further, an a priori analysis of each case for potential eCPR eligibility was conducted. A current eCPR selection criteria (from the 2-CHEER study) was used to determine eligibly.
Results

In the study window 248 eligible cardiac arrest cases were included in the OHCA registry. 30-day survival was 23.4% (n=58). The mean age of survivors was 55.4 years. 17 (6.8%) cases were deemed true refractory arrests and fulfilled the 2-CHEER eligibility criteria. The majority of these cases presented within “office hours” and no case obtained a return of spontaneous circulation standard advanced life support.
Conclusions

In this contemporary OHCA registry a significant number of refractory cases were deemed potential eCPR candidates reflecting a need for future interdisciplinary work to support delivery of this therapy.


Results
In the study window 248 eligible cardiac arrest cases were included in the OHCA registry. 30-day survival was 23.4% (n=58). The mean age of survivors was 55.4 years. 17 (6.8%) cases were deemed true refractory arrests and ful lled the 2-CHEER eligibility criteria. The majority of these cases presented within "o ce hours" and no case obtained a return of spontaneous circulation standard advanced life support.

Conclusions
In this contemporary OHCA registry a signi cant number of refractory cases were deemed potential eCPR candidates re ecting a need for future interdisciplinary work to support delivery of this therapy.

Background
Survival following out of hospital cardiac arrest (OHCA) has been historically poor, with limited improvements in outcomes over time reported in many settings. (1,2) A limited number of jurisdictions, such as King County in Washington State, have reported step-wise improvements in OHCA outcomes as the result of coordinated interagency and public health initiatives. (3,4) Rates of achieving return of spontaneous circulation (ROSC) following advanced life support (ALS) in a prehospital setting vary depending on local geography, public policy, bystander actions and logistical factors and access to healthcare resources. (5,6,7) In spite of high quality cardiopulmonary resuscitation (CPR) and best practice ALS, a proportion of OHCA cases are transported to hospital without ROSC and are termed 'refractory'. In these refractory cases, extracorporeal membrane oxygenation (ECMO), or "eCPR'' may be a management option as a bridge to de nitive therapy. (8,9) In carefully selected OHCA populations observed improvements in patient outcomes when compared to standard ALS protocols have been dramatic with multiple trials showing eCPR for OHCA and refractory ventricular brillation (VF) signi cantly improved survival to hospital discharge. (10)(11)(12)(13).
While mechanical CPR (m-CPR) does not show stand-alone survival bene t in randomised trials (14), it is most likely non-inferior to manual chest compressions and allows for effective on-going resuscitation during transport. Use of M-CPR, therefore may help facilitate earlier cannulation of potentially viable eCPR patients on arrival to hospital. (6,15,16). Implementing these systematic changes requires coordination from all disciplines of critical care and a larger whole of hospital approach to ensure appropriate clinician support and resource allocation. Furthermore, coordination with prehospital services is key in ensuring appropriate patient selection and noti cation for all potential ECMO cases.
In order to further the knowledge and understanding of the application of advanced OHCA treatment strategies in an Australasian setting, we set out to prospectively assess the baseline and future management of a contemporary OHCA cohort in a local health district (LHD) with a historically limited use of ECMO for in-patients and no history of use of eCPR in the ED.(1) Using this prospective OHCA database we applied an evidence based selection criteria to determine caseloads that might be experienced by future eCPR teams.

Study Setting and Eligibility Criteria
This project was conducted at Western Sydney Local Health District (WSLHD) in Sydney, Australia. The tertiary centre in this network, Westmead Hospital, is an urban, university-a liated hospital with a prepandemic ED census of 79,000 annual presentations. Protocols were approved by the WSLHD Human Research Ethics Committee (HREC Code 5529). The prospective observational database used for this study included consecutive ED OHCA presentations from 2016-2019. Utstein reporting methods were used for the data collection. (17) The data points were collected by a single investigator (PC) with assistance from trained data managers.
The a priori inclusion criteria were: (1) OHCA age≥16 years; (2) eMR notes accessible; (3) OHCA or cardiac arrest in ED within 1-hour of arrival. Exclusion criteria were (1) In-patient cardiac arrests (de ned as cardiac arrest>1-hour following presentation to ED); (2) Paediatric cases age<16 years. These criteria were applied to ED Electronic Medical Record (eMR) presentation lists (generated on a weekly basis over a 3-year period by an ED data manager). Eligible OHCA were manually evaluated using paper forms with additional data points obtained from eMR and state ambulance records. eCPR eligibility criteria ( Fig. 1) used by the 2-CHEER study were used to retrospectively match the likely use of eCPR in the cohort of OHCA patients (Fig. 2). (1,14) Analysis Plan

Page 4/15
The primary outcome measure (Fig. 2) was a manual point estimate of OHCA cases matching eCPR eligibility criteria (Fig. 1). For this outcome, cases were matched retrospectively by two reviewers using an a priori plan. Discrepancies in group allocation between reviewers were resolved by consensus after further review. We de ned a refractory arrest as a patient receiving CPR on arrival to the ED with an arrest to ROSC time or an arrest to death time of ≥20 minutes.(18) Secondary outcomes included demographic characteristics based on modi ed Utstein methods, 30-day all-cause mortality, assessment of factors affecting the timing of ROSC. (17) Hospital intervention including cardiac catheterisation and use of M-CPR were also recorded prospectively. Means, standard deviations (SD), medians and interquartile ranges (IQR) are used to report descriptive data as appropriate for parametric and non-parametric distributions respectively and Chi-squared tests were used for signi cance tests of reported comparative outcome data.

Results
Within the study window there were 193,750 ED presentations and 251 OHCA cases identi ed. In emergency department patient arrests (n=3) were excluded from the analysis (Fig. 2   The reported literature re ects a growing use of ECPR in conjunction with an expansion in clinical indications globally but transitioning from standard ALS to ECPR is likely to be a challenge to existing resources. (25) In addition to skilled practitioners and training the use of hospital wide and intensive care unit (ICU) resources will be required, including management of reported complications including cannulation related complications, limb ischemia, acute kidney injury, hemorrhage and stroke. (26) It is conceivable that as awareness of ECPR increases and pre-hospital MCPR is implemented that a larger number of eCPR eligible cases may present to the ED along with a possible increase of eCPR ineligible patients as well. (27) The systems level implications of a structured MCPR to ECPR process is yet to be elucidated. Furthermore, the most appropriate model to maximize "reach" of the service whilst balancing cost effectives and resource capabilities remains unclear. International experience has demonstrated the emergence of a variety of models of care around the provision of a cost effective ECMO services involving multidisciplinary training within appropriate clinical spaces which demonstrates improvements in patient care and scal outcomes.  (14), albeit without systems level costing. However, this gure falls within the commonly accepted thresholds cited above and demonstrates that ECPR may be cost effective from a scal standpoint.
Pre-hospital prognostication and management of OHCA is often challenging.(31) Although evidence is available on how to prognosticate OHCA patients, no speci c factor universally predicts survival. Therefore, there is controversy on which patients should be eligible for ECPR studies. Regarding the role of prehospital services three speci c parameters were assessed in this study including time from arrest to ROSC, time from arrest to time to triage and mean time from arrest to time of the patient being declared deceased. Our collected data for reported mean times of arrest to ROSC and mean transport times is re ective of our excellent baseline prehospital service capability and demonstrates potential to further integrate care with hospital teams. To this end further optimising transportation to hospital with ongoing ALS in refractory cases is required in order to deliver ECPR within current eligibility guidelines. Working with our ambulance colleagues in this area is a cornerstone of delivering a truly interdisciplinary and holistic OHCA response. This interface is particularly pertinent when considering the intervention of ECPR which in current guidelines includes the prehospital initiation of MCPR, an element that was not reviewed in our data. The future availability of this technology and the extent to which it is adopted by ambulance services would signi cantly impact the viability of an ECPR service.
Additionally, in this study we assessed Utstein-based OHCA outcomes in a contemporary cohort of Australian patients. The key descriptive data analysed was in keeping with the reported literature. (3) We observed that survival rates were higher in patients presenting shockable rhythms, witnessed arrests and with bystander CPR. Further it was observed that survival was higher compared to local and historical reporting of OHCA outcomes. This unexpected nding may be anomalous due to a selection bias in this ED cohort associated with paramedic protocols allowing CPR cessation in the eld. (19) Of particular interest is the importance of immediate effective bystander CPR, re ected by the survival outcomes in this study without bystander CPR (16.6%) versus with bystander CPR (30.5%). Future local work must focus on strategies to reduce the apparent high proportion of patients who do not receive bystander rescue (27.8%, n=65). A recent relevant meta-analyses suggested community based interventions are e cacious in improving rates of bystander CPR. (20) Population surveys indicated only 22% of the Australian public are trained in CPR provision, with these numbers remaining unchanged over the last decade (32), and is low by international standards. Concerted education and training programs have shown signi cant bene ts in improving bystander CPR rates (Singapore lancet article). Local efforts have begun that will hopefully provide improvements. (21) The likelihood of initiating high-quality CPR has been shown to increase even after very brief community-based education programs focused on basic life support. (22) Improving the quality of bystander and prehospital care of OHCA victims holds the key to driving incremental improvement in survival as well as increasing the number of patients eligible for eCPR in the event of a refractory OHCA. Public-health based approaches should be viewed as the cornerstone of an integrated approach to improving resuscitation outcomes. (3,23) Limitations The major limitations of this study relate to the speci c single-centre urban location limiting generalisability. Furthermore case load estimation is obtunded by the exclusion of patients in whom resuscitation efforts were ceased in the eld due to a lack of ROSC. It is unclear how many of these patients would potentially be included in the context of the prehospital M CPR use. However, given that the case load estimation was performed in a system without the current use of prehospital M-CPR the inclusion of these patients would be as yet untenable within the bounds of the current system. They do however represent a mythological aw within the derived data set. In addition observational studies are associated with inherent biases and the data extraction in the study was from multiple sources (prehospital, ED and ICU). Therein, it is possible that clerical errors in the original data entry may in turn have led to inaccuracies in our results. Improving OHCA outcomes requires not only a co-ordinated multidisciplinary approach within hospital-based models of care, but also requires liaison with all stakeholders and services.
Two potential limitations pertain to the methodological aws of this study and possibly effect the estimation of potential caseload numbers. The limitation of excluding non-transported patients and the single centre nature of the study potentially missing patients who were transported to other facilities within the area health service presents a selection bias that likely would cause an underestimation of true case numbers. The implementation of M-CPR into the prehospital service provision may cause an increase in the number of presentations for ECMO applicable patients and may in fact cause an increase in the presentation of patients in general to centres that offer ECMO services. Appropriate inclusion criteria, the provision of training and patient support resources (social workers etc.) and the use of diagnostic investigations in patients on ECMO all require resource allocation and focus as part of the the planning from a systems perspective. (33)

Conclusion
The results re ect a higher than traditionally expected OHCA survival rate in a contemporary cohort of ED patients. At baseline more than 5% of OHCA patients may be eligible for eCPR based on the 2-CHEER criteria. The demonstrated low survival in those patients deemed to have "refractory" cardiac arrests further attests to the potential bene ts that may be gained from the implementation and use of eCPR in a coordinated and well-resourced system and in conjunction with public health education to improve rates and quality of bystander CPR in the community. Further studies and cost bene t analysis are needed to identify where future local quality improvement strategies should be focused. Out of Hospital Cardiac Arrest (OHCA) STARD Diagram