Article Text
PDF
Abstract
Background National data collection provides information on out-of-hospital cardiac arrest (OHCA) incidence, management and outcomes that may not be generalisable from smaller studies. This retrospective cohort study describes the first 2 years' results from the Irish National Out-of-Hospital Cardiac Arrest Register (OHCAR).
Methods Data on OHCAs attended by emergency medical services (EMS) where resuscitation was attempted (EMS-treated) were collected from ambulance services and entered onto OHCAR. Descriptive analysis of the study population was performed, and regression analysis was performed on the subgroup of adult patients with a bystander-witnessed event of presumed cardiac aetiology and an initial shockable rhythm (Utstein group).
Results 3701 EMS-treated OHCAs were recorded for the study period (1 January 2012–31 December 2013). Incidence was 39/100 000 population/year. In the Utstein group (n=577), compared with the overall group, there was a higher proportion of male patients, public event location, bystander cardiopulmonary resuscitation (CPR) and early defibrillation. Median EMS call–response interval was similar in both groups. A higher proportion of patients in the Utstein group achieved return of spontaneous circulation (35% vs 17%) and survival to hospital discharge (22% vs 6%). After multivariate adjustment for the Utstein group, the following variables were found to be independent predictors of the outcome survival to hospital discharge: public event location (OR 3.1 (95% CI 1.9 to 5.0)); bystander CPR (2.4 (95% CI 1.2 to 4.9)); EMS response of 8 min or less (2.2 (95% CI 1.3 to 3.6)).
Conclusions This study highlights the role of nationwide registries in quantifying, monitoring and benchmarking OHCA incidence and outcome, providing baseline data upon which service improvement effects can be measured.
- prehospital care
- resuscitation
- cardiac arrest
Statistics from Altmetric.com
Key messages
What is already known on this subject?
Data from out-of-hospital cardiac arrest (OHCA) studies that are carried out in small populations or single communities may not be generalisable to entire countries.
Comprehensive national data collection is required for monitoring of nationwide OHCA incidence management and outcomes.
What might this study add?
Our study describes Irish OHCA incidence and outcomes and shows that nationwide OHCA data collection in a population of 4.6 million is feasible and sustainable.
Introduction
While studies of single communities provide data on out-of-hospital cardiac arrest (OHCA) outcome, nationally representative data are essential in monitoring national trends in OHCA survival.1 In Europe and North America, over 575 000 cases of OHCA occur annually, and the clinical and societal impact of OHCA is such that the American Heart Association has recommended that it be classified as a reportable disease.2 In the absence of rapid recognition, good-quality cardiopulmonary resuscitation (CPR) and early defibrillation, there is a negligible prospect of survival from OHCA.3
Reported incidence of and outcomes from OHCA vary widely internationally. The reasons for this variation include: variation in age and gender distribution; levels of urbanisation; bystander CPR; availability of community defibrillators; and configuration of emergency medical services (EMS).4–6 Intercountry differences are affected by these factors and are also complicated by variation in study sample definition and denominator.7 This is despite the widespread acceptance and use of the Utstein criteria for OHCA data collection.8
The Irish National Out-of-Hospital Cardiac Arrest Register (OHCAR) was established in 2007 to provide data to estimate the incidence and survival of OHCA in Ireland, with the aspiration of improving it. OHCAR is funded by the National Ambulance Service and Prehospital Emergency Care Council. It is administered and academically supported by the National University of Ireland, Galway and hosted by the Department of Public Health Medicine in the Irish Health Service Executive. In 2012, OHCAR achieved comprehensive national data collection.9
Within this context, the aim of this paper is to provide an overview of the first 2 years of comprehensive Irish data collection and describe the incidence, key interventions and survival outcomes for OHCA attended by ambulance services where resuscitation was attempted (EMS-treated OHCA) in Ireland. In line with the Utstein guidelines and as recommended by Chamberlain and Eisenberg, we focus our analysis on the subgroup of adult patients, with presumed cardiac aetiology, with a bystander-witnessed event and an initial shockable rhythm (Utstein group).10
Methods
In Ireland, 62% of the total population of 4.6 million lives on 2.4% of the total land area.11 The remaining population is dispersed in low-density settlements across the country. In Dublin city and county, the National Ambulance Service (NAS) and the Dublin Fire Brigade (DFB) provide the statutory EMS response, while throughout the rest of the country, statutory ambulance services are solely provided by the NAS. All prehospital emergency practitioners who work for statutory ambulance services must be registered with the Pre-Hospital Emergency Care Council. Practitioners use clinical practice guidelines to inform decisions not to resuscitate or cease resuscitation.12 Emergency medical technicians and emergency first responders are trained in basic life support, including automated external defibrillator (AED) use. Paramedics can perform supraglottic airway placement and advanced paramedics are additionally trained to intubate in cardiac arrest situations, attempt manual defibrillation and administer cardiac resuscitation drugs. All ambulance vehicles are staffed with paramedics and/or advanced paramedics. For cardiac arrest calls, the DFB also deploy practitioners on fire engines. Community response to OHCA in Ireland varies. In some areas, the community response depends on the training and willingness of people to perform basic life support and the opportunistic availability of AEDs in the vicinity of the event. In other areas, the level of response is highly organised and coordinated by voluntary Community First Responder programmes. Irish general practitioners, primarily in some rural areas and some county fire services, also respond to OHCA at the request of the NAS.13
Statutory ambulance services in Ireland use a standardised Patient Care Report (PCR) which includes an ‘OHCA’ section for Utstein required data. For incidents attended by the DFB, all PCRs are received at a central location where PCRs for OHCAR incidents are manually identified. Data from each OHCAR incident are entered onto a Microsoft Access database. Each incident is electronically linked to corresponding dispatch data, and completed records are sent to OHCAR on a quarterly basis. Data are then checked to ensure compliance with OHCAR definitions as well as to avoid double entries.
In the rest of the country, immediately after attending an OHCA incident, NAS practitioners put completed PCRs for OHCAR incidents in specially provided envelopes. These PCRs are digitally processed at a central location and then electronically forwarded to OHCAR for case-by-case validation. Dispatch data are then added to all cases. In order to identify cases that may not have been placed in envelopes, missing case searches are performed in the NAS digital PCR archive. Outcome data are also obtained for patients brought to hospital.
Data for this study were anonymised and extracted from the OHCAR database for the study period.
Ethical approval
Ethical approval for research using non-identifiable OHCAR data was obtained from the Research Ethics Committee, National University of Ireland, Galway.
Statistical analysis
This is a retrospective cohort study of incidence, interventions and outcomes of EMS-treated OHCA in Ireland during 2012 and 2013. The reported population for 2012 and 2013 is described. To allow international comparisons and be in line with Utstein recommendations, a subgroup of patients was extracted, which includes only adult patients with a bystander-witnessed OHCA, of presumed cardiac aetiology and an initial shockable rhythm (Utstein group).10 To ensure our data can be compared with data from other national registries, incidences per 100 000 population per year for the total group and the Utstein subgroup were standardised for age and sex using the 2013 EUROSTAT population projections. In order to describe the difference in OHCA incidence according to age, age-adjusted incidence for the total group and Utstein subgroup is graphically presented. The key outcomes from our analysis are to calculate survival to discharge for the overall group and for the Utstein group, and to determine predictors of survival in the Utstein group.
An overview of differences between the total group and the Utstein group is presented. A variable was derived to represent the availability of early defibrillation using the following rule: (defibrillation attempted=yes AND (EMS response interval of 5 min or less OR defibrillation attempted before EMS arrival)). Analysis of predictors of survival was limited to the Utstein group only.
Logistic regression analysis was performed to identify predictors of the main outcome of interest, that is, survival to hospital discharge. Variables were entered into the model based on: at least moderate univariate associations (p<0.15); validation of significance in previous literature; clinical relevance to support inclusion. Continuous variables, that is, age and ambulance response times, were categorised for regression analysis. Calibration of the model was assessed using the Hosmer and Lemeshow χ2 statistic (p>0.05). In order to assess the potential effect of loss to follow-up, two potential scenarios were created: (i) assumed all missing cases had survived to hospital discharge; (ii) assumed all missing cases had died. Logistic regression analysis was repeated for both scenarios. Description and analysis of all cases of non-traumatic aetiology were also performed (see online supplementary tables S1 and S2).
Supplementary tables
Results
A total of 3701 EMS-treated OHCAs were recorded for the study period (1798 in 2012 and 1903 in 2013). The overall incidence of EMS-treated OHCA was 39/100 000 population/year. The Utstein group had an incidence of 6/100 000 population/year. Age-adjusted incidence was highest for overall cases in the 85+ age group, but peaked in the 70–74 year age category for the Utstein group (figure 1).
Age and sex adjusted incidence per 100 000 population per year (EUROSTAT, 2013).
For the overall group, 855 cases were missing one or more descriptive variables (23%), including 30 patients who were lost to follow-up. In the Utstein group, nine patients were lost to follow-up. As shown in table 1, 6% of all patients survived to hospital discharge, compared with 22% of patients in the Utstein group. Median age for all cases was 67 years (IQR 52–78 years), with the majority of patients aged over 65 years. Over half of patients (54%) had a bystander-witnessed arrest, and 70% of these patients received bystander CPR. Most cases were presumed to be of a cardiac aetiology (86%). Trauma (including self-harm and road traffic accidents) accounted for 7% of cases. Other causes included submersion and drug or alcohol overdose. The Utstein group comprised 15% of all cases (n=577). Patients in this subgroup were similar in age and gender to the overall group, but had higher percentage survival to discharge (22%) and better secondary outcomes, that is, return of spontaneous circulation, than the overall group. The Utstein group also had higher proportions of publicly located events, bystander CPR and defibrillation attempted. The Utstein group also had, as shown in figure 2, a higher percentage of EMS call response within 8 min, and this marginal difference in proportions persisted for the majority of response intervals.
Overview of the differences in descriptive variables and outcome between all cases and the Utstein group
Cumulative percentage of cases responded to at each emergency medical services (EMS) call–response interval.* *Time in minutes from call pick-up in ambulance control centre to first EMS vehicle arrival at scene. EMS-witnessed cases excluded.
As shown in figure 3, complete data were available for 502 of the 577 patients in the Utstein group. As shown in table 2, age under 65, collapse in an urban setting, collapse in a public location, bystander CPR, early defibrillation attempted and an EMS response of 8 min or less were all associated with patients' survival to discharge in the univariate analysis (model 1). In the logistic regression model (model 2), public location of the OHCA incident (OR 3.1 (1.9 to 5.0)), bystander CPR (OR 2.4 (95% CI 1.2 to 5.0)) and EMS response time of 8 min or less (OR 2.2 (1.3 to 3.6)) remained significant predictors of survival to discharge. Interactions between variables were not significant and omitted from the model. Data on outcome were missing for nine patients in the Utstein group. The analysis was repeated using the assumption that (1) all missing cases had survived or (2) all missing cases had died. In both models, adjusted ORs for all variables remained similar.
Regression analysis for Utstein group for the outcome survival to discharge
Cases included in logistic regression analysis. CPR, cardiopulmonary resuscitation; ROSC, return of spontaneous circulation.
Discussion
Ireland is one of the few countries in Europe where nationwide reporting of OHCA is currently possible and a system has been developed that allows routinely collected data to be used to build a national register of OHCA capable of providing meaningful risk-adjusted quality indicators. This paper provides a description of EMS-treated patients with OHCA and their outcome in Ireland during 2012 and 2013. Our paper highlights the value of quality registries in describing, benchmarking and highlighting where challenges arise in care delivery and solutions hypothesised.
The incidence of OHCAR was 39/100 000 population/year. There is wide variation internationally in the reported incidence of OHCA, from 19 to 141 per 100 000 population per year,7 suggesting variation in the threshold to commence CPR. The total number of incidents was similar in 2012 and 2013, suggesting internal consistency in case identification and incidence. The proportion of incidents that occurred in urban areas reflects the proportion of the Irish population that resides in urban areas. The age and gender profile of patients was very similar to other large studies of OHCA, as was the difference in median male and female ages.4 ,14 The proportion of patients presumed to have suffered an arrest of cardiac aetiology is similar to North American data, but high compared with other population-based studies of EMS-treated OHCA.11 ,14 This may be explained by the fact that in Ireland cardiac aetiology is presumed in the absence of documented evidence of any other probable cause and may be considered analogous to ‘unknown cause’.
While the proportion of patients in our study with an initial shockable rhythm was 24%, there is significant variation in population-based studies, from 8.7% in Japan to 36% in North America.14–16
Reported percentage survival to discharge globally varies from 0.8% to 25% in OHCA attended by ambulance services. At 6%, percentage survival in Ireland is low, but many other studies are region-specific and do not reflect national survival. While there is considerable heterogeneity in the overall group, the Utstein subgroup includes patients who have been proven to benefit most from a resuscitation attempt, that is, adults, cardiac cause, bystander-witnessed and initial shockable rhythm.10 Survival in the Utstein subgroup was 22%, which is substantially lower than in other population-based studies where percentage survival of up to 52% has been achieved.17 ,18 Opportunities to strengthen the chain of survival in Ireland are being vigorously pursued.
In this study, we found that collapse in a public location, provision of bystander CPR and an EMS call–response interval (CRI) of 8 min or less were independent predictors of survival to hospital discharge. Our finding that collapse in a public location accounted for a threefold increase in survival is not surprising and has been reported in other large-scale registry-based studies.19 ,20
The proportion of bystander CPR provided in bystander-witnessed cases was high at 70% and even more impressive in the Utstein subgroup where bystander CPR was an independent predictor of survival. The proportion of bystander CPR provided is similar to the countries and regions where percentage survival is higher than that in Ireland.21 ,22 Reliability of bystander CPR measurement is an issue for all OHCA registers, but we believe the nationwide introduction of dispatch-assisted CPR may help account for the high levels of bystander CPR observed in our study. Wissenberg et al21 observed an association between improved OHCA survival in Denmark and increased bystander CPR rates following national initiatives to increase bystander interventions. Our results indicate willingness among the Irish population to attempt CPR and suggest that further extension of training initiatives may contribute to improve survival rates in Ireland.
An EMS CRI of 8 min or less was also an independent predictor of survival. Identifying ways in which to minimise this interval is essential to improving Irish OHCA outcomes. In 2015, the NAS introduced ‘ONELIFE’, an extensive programme to improve OHCA outcomes within the NAS.23 As part of ONELIFE, dynamic deployment of EMS vehicles has been implemented, and strategies to improve dispatcher OHCA recognition and incident location are currently being introduced.
For this study, we derived a variable to represent early defibrillation and found that a minority of patients had access to early defibrillation. We had expected that early defibrillation would be a significant predictor of survival in multivariate analysis. The lack of significance may be because the derived variable overestimated the availability of early defibrillation, as we assumed that all defibrillation attempts before EMS arrival were made within minutes of collapse. This may not have been the case, particularly in more rural areas where travel times, even for first responders, may be prolonged. Blom et al24 described how AED use was an independent predictor of survival in the Netherlands. In their study area, AED use was tripled as a result of policy measures, including: introduction of AED programmes for police teams (together with existing fire service response); implementation of a ‘6-min time zone’ and the introduction of a text alert system for registered volunteers. Structured AED programmes also have the advantage of efficiency as described by Ringh et al.22 They reported similar percentage survival as a result of 74 deployments from 5016 public AEDs compared with 53 deployments out of a possible 135 first responder AEDs, suggesting that coordinated support of first responder programmes would be more cost-efficient than mass implementation of public AEDs. In Ireland, voluntary groups, general practitioners and county fire services already provide a community response to OHCA. Extension and support of such schemes is considered an important way in which to reduce time to defibrillation.
Limitations
A substantial percentage of data was missing for the overall group, most notably resulting in 75 cases being omitted from the logistic regression analysis. In order to assess the impact of missing data, missing data imputation was performed for all cases, and logistic regression analysis was repeated for the Utstein group. The pooled results from regression analysis using imputed data did not differ significantly from the results found using original data (see online supplementary table S3).
Thirty patients were lost to follow-up. Most of these patients could not be traced due to unavailability of patient identification or poor legibility of PCRs. In our study, loss to follow-up did not significantly affect results, but it remains an issue for OHCAR.
We presumed a cardiac cause in over 86% of OHCAR cases. This presumption may have led to misclassification bias.25 Classification of cases as ‘presumed cardiac’ was originally proposed by the Utstein Committee to create ‘case equivalency’; however, such classification can be highly subjective.8 Reporting of EMS-treated cases of non-traumatic aetiology is a suggested way in which to decrease subjectivity and improve comparability of registries worldwide (please see online supplementary data).
Conclusions
This study provides a nationwide description of EMS-treated OHCA in Ireland. The incidence and demography of OHCA is similar to other population-based studies. Initiatives to increase public education in CPR, support further implementation of community first responder programmes and continued quality improvement in the EMS are keys to improving OHCA outcomes. This nationwide profile provides the dashboard by which improvements can be measured.
Acknowledgments
The authors wish to thank National Ambulance Service and Dublin Fire Brigade personnel who provided the clinical and dispatch data that have made this study possible.
References
Footnotes
Contributors SM: performed data analysis and drafted the text of the manuscript. JC: co-supervised. SM: wrote, read and reviewed manuscript drafts. CD: contributed to the introduction and discussion text. MO: contributed to the methods section. PW: contributed to the discussion section. AM: contributed to the discussion and conclusion sections. BM: reviewed text and advised on data analysis. AV: co-supervised. SM: drafted the text, contributed to manuscript text, instructed on data analysis plan and checked results.
Funding SM is funded by the Health Professionals Fellowship Award from the Health Research Board, Ireland (HPF-2014-609).
Competing interests BM reports grants from American Red Cross, the American Heart Association, Medtronic Philanthropy and Zoll Corporation, outside the submitted work.
Ethics approval Research Ethics Committee, NUI Galway.
Provenance and peer review Not commissioned; externally peer reviewed.