Abstract
Keywords
Introduction
Out-of-hospital cardiac arrest (OHCA) is a time-critical emergency that imposes significant healthcare challenges and public health burdens worldwide. 1 Timely recognition of the emergency and prompt activation of the emergency system by bystanders are the critical starting points in the chain of survival following an OHCA. Early recognition of cardiac arrest and initiation of cardiopulmonary resuscitation (CPR) is imperative to successful resuscitation and favourable outcomes.2–5 A meta-analysis that involved 232,703 patients showed that bystander CPR was associated with an improved OHCA survival (odds ratio = 1.95), especially in patients with a shockable rhythm. 2 It is important to increase the bystander CPR rate in OHCA to improve survival.
In Hong Kong, the annual incidence of OHCA was 72 per 100,000 population in 2012. The survival to hospital discharge (STD) rate was 2.3% and only 1.5% of the survivors had a favourable neurological outcome, 6 which was among the lowest in Asia. 7 Notably, bystander CPR was performed only in 28.8% of the OHCA cases in Hong Kong. 6 Local survey revealed poor knowledge of CPR and a lack of relevant training in the general public. 8
The provision of dispatch-assisted cardiopulmonary resuscitation (DA-CPR), also known as telecommunicator CPR (T-CPR), aims to augment the survival rate of OHCA by empowering lay persons to provide timely resuscitation to OHCA victims before the arrival of an ambulance under the guidance of call-takers (or dispatchers) of the emergency call centre. 9 In Hong Kong, the Hong Kong Fire Services Department (HKFSD) has implemented the DA-CPR under the post-dispatch advice (PDA) service to enhance its emergency ambulance service (EAS) since October 2018. However, the local impact of PDA on the bystander CPR rate and OHCA survival in Hong Kong is not clear. Also, the reasons for caller refusal of PDA on CPR have not been evaluated in detail locally.
This study was conducted to address the above knowledge gaps. We aimed to (1) evaluate the impact of PDA on the bystander CPR rate and OHCA survival, hypothesising that PDA increases in Hong Kong, and (2) characterise the reasons for not providing bystander CPR when PDA is offered, which have important implications in public CPR education and PDA service improvement.
Methods
This was a territory-wide retrospective review of all OHCA-related calls to the Fire Services Control Center (FSCC) and the EAS responses from the HKFSD between 1 September 2021 and 31 October 2021. The study methods, results, and discussion are reported according to the Strengthening the Reporting of Observational studies in Epidemiology (STROBE) guidelines. 10
Study setting
In Hong Kong, most OHCA victims are first attended by the EAS of the HKFSD. All EAS are activated by a unique emergency number (call number 9-9-9) connected indirectly through police call centre to the emergency dispatching centre (FSCC), which is staffed by around 200 dispatchers. The FSCC handles about 2000–2200 calls daily, totalling 700,000 calls per year. The EAS provided by the HKFSD is a one-tiered system serving the entire 1104 km2 of Hong Kong and provides 24/7 emergency services to a population of approximately 7.4 million residents during the study period.
Since October 2018, the HKFSD has employed the Medical Priority Dispatch System (MPDS), developed and franchised by the International Academies of Emergency Dispatch (IAED) at Salt Lake City in the United States, in the FSCC. The dispatchers in the FSCC are non-medical professionals who have undergone mandatory initial certificate training courses to use the system with continuous dispatch education modulated by the IAED. The new system enables dispatchers of the FSCC to provide callers with immediate, comprehensive and protocol-driven PDA on more than 30 types of injuries and illnesses (including physical trauma, loss of consciousness and cardiac arrest) after dispatching the ambulances.
The protocols incorporated in the computer system were developed by the IAED with continuous quality improvement. Similar computer systems and corresponding protocols have been adopted by over 3000 mobilising centres of emergency services in more than 40 regions, including Mainland China, the United States, Canada, the United Kingdom, France, Australia, and so on. All FSCC personnel have been trained to become emergency medical dispatchers. A quality improvement unit has been set up in the HKFSD to ensure satisfactory delivery of PDA through regular audit of a portion of calls and provision of feedback to the dispatchers.
In the case of a suspected cardiac arrest, the dispatcher provides PDA following two key questions: (1) ‘Is the victim conscious?’ and (2) ‘Is the victim breathing normally?’. Conventionally, If the caller answers ‘no’ to both questions, instruction for CPR is indicated (Figure 1). This ‘no-no-go’ approach might miss some cardiac arrest cases in the dispatch environment. 11 In the FSCC of HKFSD, identification of cardiac arrest is further enhanced by additional protocol-driven telephone interrogations, they include and not limited to: (1) regard no breathing if the breathing status cannot be provided by the caller at scene; (2) dispatchers are encouraged to use an ‘agonal breathing diagnostic tool’ whenever in doubt; (3) dispatchers are encouraged to stay online with the callers if the patient is categorised as unstable by the protocol; and (4) dispatchers are empowered to start advising CPR in any doubt. In the local setting, compression only CPR was advised for most non-traumatic cardiac arrest cases in adults. Chest compression to ventilation in 30:2 ratio was advised in paediatric patients, trauma, poisoning and airway obstruction cases until the arrival of the EAS personnel from the HKFSD at the scene. Automated external defibrillator (AED) instruction can also be given by dispatcher if available at the scene. Meanwhile, the computer system will automatically send the patient information to the responding ambulance to enable the responding crew to make better preparation before scene arrival. The EAS personnel are required to perform CPR and transfer most cases of OHCA to emergency departments (EDs) of public hospitals under the hospital authority (HA). A small proportion of patients with obvious post-mortem changes such as rigor mortis are directly transferred to public mortuary.
The PDA algorithm.
Study population
Data of all adult patients (⩾18 years) with OHCA attended by the EAS of the HKFSD who were resuscitated and transferred to the ED by ground ambulance during the study period were included. We excluded (1) patients who were not transferred by ground ambulance, (2) patients who were directly transferred to public mortuary, (3) patients who had been receiving bystander CPR before the emergency call and (4) patients with incomplete DA-CPR outcome data.
Data collection and definition
The data collected included the location of the cardiac arrest, whether the cardiac arrest was witnessed and the person who witnessed it, performance of bystander CPR and the persons involved, delivery of prehospital defibrillation with an AED, the first cardiac rhythm detected by the EAS and the time of prehospital events (recognition of cardiac arrest, EAS call receipt, start of bystander CPR, first defibrillation, EAS arrival at patient’s side, start of CPR by EAS personnel and ED arrival). Most data elements for OHCA were defined according to the Utstein Template guidelines. 12
We reviewed the ED chart and electronic medical records in the HA Clinical Management System (CMS) and followed all included OHCA till death or hospital discharge. The primary outcome was STD. The secondary outcomes included return of spontaneous circulation (ROSC), either before or after ED arrival, and survival to hospital admission (STA).
Bystander CPR was defined as any attempt at chest compressions, with or without ventilation, by a person who was not responding as part of the organised emergency response system to a cardiac arrest. 12 The occurrence of bystander CPR was determined based on review of all PDA audio recordings by two investigators independently to assess whether PDA was given to assist bystander CPR and the callers’ response to PDA on CPR. By listening to the PDA audio recordings, bystander CPR was regarded as present if the callers did not refuse and followed instructions given by the dispatcher, meanwhile the dispatcher would stay on the phone conversation and count the CPR rhythm with the callers until the arrival of Emergency Medical Service (EMS) personnel. The reasons for refusal were reviewed through a thematic analysis under the following categories: caller’s proximity to the patient, delayed recognition of OHCA, physical limitations, language barriers, emotional state, connection issue and personal perceptions. Consensus was reached by discussion if there were discrepancies.
Statistical analysis
We did not calculate the sample size before the study because there was a lack of local study to guide such a calculation. We did not impute missing values and cases with incomplete outcome data were excluded from analysis. Descriptive statistics was used to analyse the distribution of categorical variables (counts and proportion) and continuous variables (medians and inter-quartiles). We compared the distribution of clinical and outcome characteristics between groups of patients who received PDA bystander CPR and those did not. Chi-square test or Fisher exact test was used to compare categorical variables and Mann–Whitney U test for continuous variables between groups, as appropriate. We performed subgroup analysis based on the patient’s age (⩽65 vs >65 years), whether the cardiac arrest was witnessed and whether the first detected cardiac rhythm was shockable. Significance level was set at a p-value of <0.05. Data analysis was performed using the Statistical Package for the Social Sciences (SPSS) for Windows version 27.0 (IBM Corp., Armonk, NY, USA).
Ethics statements
The study was approved by the HKFSD and HA Hong Kong East Cluster Research Ethics Committee (IRB no: IRB 2022-016). The requirement for patient consent was waived by the ethics committee because of its retrospective design and anonymity in data analysis.
Results
A total of 1026 OHCAs were attended by the EAS during the study period, of whom 219 were excluded due to incomplete OHCA outcome data and 58 were excluded because of bystander CPR initiation before the emergency call. We included 749 OHCA cases for analysis (Figure 2).
Patient flow diagram.
The patient demographic, clinical and outcome characteristics are shown in Table 1. The median age of the patient was 82.0 years (interquartile range (IQR) = 67.0–90.0 years) and 55.1% were men. The overall DA-CPR rate was 49.6%, with 256 (68.8%), 76 (20.4%) and 39 (10.5%) OHCA victims who received bystander CPR in home, elderly home, public or workplace, respectively. Only 10 OHCA patients received bystander public access defibrillation (PAD).
The demographic, clinical and outcome characteristics of out-of-hospital cardiac arrest cases given post-dispatch advice.
CPR: cardiopulmonary resuscitation; PDA: post-dispatch advice; IQR: interquartile range; PAD: public access defibrillation; AED: automated external defibrillation; ROSC: return of spontaneous circulation; STA: survival to hospital admission; STD: survival to hospital discharge.
Mann–Whitney U test.
Chi-square test.
Patients who received DA-CPR had a higher proportion of witnessed cardiac arrest (37.7% vs 24.7%, p < 0.001) and a shorter time interval from recognition of cardiac arrest to chest compression (median 3.0 min vs 13.0 min, p < 0.001) compared with patients without DA-CPR. The two groups did not differ significantly in terms of age, sex, location of cardiac arrest, first rhythm and prehospital defibrillation.
Overall, 126 (16.8%) OHCA victims had ROSC before or upon arrival at the hospital. Patients who received DA-CPR had a higher ROSC (18.3% vs 15.4%), STA (15.9% vs 13.3%) and STD (2.7% vs 1.3%) rates compared with those who did not. The absolute risk reduction (ARR) is 2.9% (95% confidence interval, −2.5% to 8.3%). However, the differences did not reach statistical significance.
With subgroup analysis, patients ⩽65 years who received DA-CPR had a significantly higher ROSC rate compared with those who did not. For patients with witnessed arrest and shockable first cardiac arrest rhythm who received DA-CPR, they had a higher ROSC (30.7% vs 22.6% and 60.0% vs 39.1%), STA (25.7% vs 20.4% and 60.0% vs 34.8%) and STD (5.7% vs 2.2% and 13.3% vs 4.3%) rates when compared with those who did not receive DA-CPR, although the differences did not reach statistical significance (Table 2).
Subgroup analysis of out-of-hospital cardiac arrest survival in patients who received bystander cardiopulmonary resuscitation according to post-dispatch advice compared with those who did not.
CPR: cardiopulmonary resuscitation; PDA: post-dispatch advice; ROSC: return of spontaneous circulation; STA: survival to hospital admission; STD: survival to hospital discharge.
Chi-square test.
Fisher exact test.
Bystander CPR was not offered to 377 OHCA victims despite PDA was given, with a refusal rate of 50.3%. The reasons for refusal are summarised in Table 3. Of note, 48% of the callers were unable to perform resuscitation as they were not physically present by the patient’s side at the time of emergency call.
Summary of reasons for callers refusing cardiopulmonary resuscitation despite post-dispatch advice.
CPR: cardiopulmonary resuscitation.
6.9% of the callers were unable to perform CPR due to various technical difficulties (e.g. concerns about scene safety, inaccessibility of the patients, inability to position patient for CPR) and 7.4% of the callers declined to perform CPR as it was deemed inappropriate by the callers due to underlying illnesses or old age of the victims.
Discussion
Our study was the first territory-wide study on OHCA survival in Hong Kong after implementation of the PDA system by the HKFSD. The data demonstrated a bystander CPR rate of 53.3% (including 58 bystander CPR initiated before the emergency call and 372 DA-CPR), which is higher than the reported bystander CPR rate of 28.8% in 2017. 6 Most bystander CPR (86.5%) was initiated after PDA instruction. The STD rate was higher in patients who had received bystander CPR, although it did not reach statistical significance.
The increase in the bystander CPR rate after implementation of PDA is consistent with previous studies. A meta-analysis on 235,550 patients, from populations of Japan, America, Sweden, Canada, Finland and Korea, showed that DA-CPR is effective in improving the bystander CPR rate. 13 Implementation of the national dispatcher-assisted CPR intervention in Singapore increased the bystander CPR rate from 22.4% to 42.1%. 14 The higher bystander CPR rate observed in the current study could be explained by the following reasons: (1) increased awareness and earlier recognition of cardiac arrest with the aid of PDA; (2) enhanced confidence and motivation of the bystanders to perform CPR with instructions and encouragement from the dispatchers; and (3) difference in the method used to capture the occurrence of bystander CPR. In contrast to the previous local study on OHCA, 6 in which the occurrence of bystander CPR was based solely on the recollection and documentation of the EAS personnel, bystander CPR was captured in the current study by reviewing individual dispatch audio recordings, which minimised the recall bias from recollection.
Despite the significantly shorter time interval from cardiac arrest recognition to initiation of chest compression in DA-CPR, the improvement in ROSC, STA and STD rates did not reach statistical significance. Yet in post hoc subgroup analysis, in those aged <65 years, the ROSC rate was significantly higher. The relatively old age of the study population may have diluted the effect of DA-CPR, which also explained the lack of significant difference in the overall survival rate. Zhang et al. 15 reported similar experience when implementing DA-CPR in Suzhou, China, and they attributed the failure to improve survival to the prolonged delay in calling the emergency services. Likewise, a low percentage of witnessed arrest in our study might indicate a significant delay from collapse to CPR, rendering the effort of DA-CPR futile. In addition, we believe other factors might affect survival. First, the quality of bystander CPR was not guaranteed. We did not have data on the CPR competency of the bystanders and CPR quality. Most OHCAs happened at home and in some cases, CPR was performed by the spouses of OHCA victims, who may be elderly bystanders with a lower physical strength, resulting in less effective CPR. 16 Also, the sample size might have been inadequate to show a significant difference in the ROSC and survival rates.
Through a thematic analysis of the OHCA dispatcher audio recordings, we identified several barriers to bystander CPR: caller’s proximity to the patient, delayed recognition of OHCA, physical limitations, language barriers and emotional state. In this study, we found that bystander CPR rate is significantly higher in patients with witnessed cardiac arrest (37.7% vs 24.7%). OHCA recognition by the callers and dispatchers is positively associated with bystander CPR 17 and ‘not breathing’ was associated with recognition of OHCA by call handlers. 18 A review of 100 audio recordings of dispatch calls in Sweden found that agonal breathing patterns hampered the callers’ willingness to perform CPR as callers often mistaken the agonal breathing as a sign of life. 19 Another study on 404 audio recordings in an American EMS system showed that bystander reporting signs of life was the most common barrier to the dispatchers not providing CPR instructions. 20 These findings highlight the importance of correct recognition of signs of life during call handling. Yet, how to discern agonal breathing from normal breathing might still be difficult for a lay person.
Studies in America and Singapore reported technical difficulty to move the patient to an appropriate position for CPR as a significant barrier, which was similarly found in our study.14,20 The caller’s high emotional state is not a major barrier to OHCA recognition by dispatchers, but it may increase the difficulty in delivering DA-CPR instruction. However, if dispatchers can skilfully reassure the emotional callers, it is still possible that callers can follow DA-CPR instruction to perform chest compression despite the emotional state.20–22 Interestingly, the barriers to DA-CPR appear to be quite different from that found for bystander CPR in general, such as fear of medicolegal consequences and concerns of disease transmission.20,23,24 One possible explanation is that the majority of OHCA happened at home (66.7%) and DA-CPR were mainly provided by their relatives.
In our study, a surprisingly high proportion (48%) of the callers declined PDA on bystander CPR because they were not physically with the victim. From listening to the audio recordings, we discovered that these callers were usually the children of the patient, who were informed by the elderly spouse or domestic helper about patient’s emergency condition remotely by phone and decided to call EMS for help. It is consistent with another study in China that in case of a medical emergency, Chinese people, especially the elders, are more likely to call their families or neighbours for help instead of calling EMS directly. 15 Those who made the emergency call were often their family members who might not be physically present by the patient side. This could be improved if the dispatchers ask for telephone contact and give PDA directly to the onsite companion of the victim. We also identified a number of callers who refused DA-CPR as they deemed CPR in the patients was not appropriate, for example, due to terminal illness, old age or advance directive. For CPR in OHCA victims with terminal illness and advance directive, ambulance personnel of the HKFSD are currently bound by the Fire Services Ordinance (‘FSO’) (Cap. 95) to perform resuscitation on any person who appears to need prompt or immediate medical attention. 25 There is yet to have clear legislative provisions on the relationship between duties under the FSO and an advance directive.
To overcome these barriers to DA-CPR and improve the outcome of OHCA victims with PDA, targeted public education is necessary to provide information about what occurs in an emergency call, how to recognise an OHCA, including abnormal breathing pattern, and to improve CPR knowledge and skills in the community. 26 Another promising solution to tackle these barriers is to dispatch citizen responders through smartphone application. The arrival of app-dispatched citizen responders before EMS was associated with increased odds for bystander CPR and bystander defibrillation. 27
Nevertheless, low survival and ROSC rate may also be explained by a low rate of bystander defibrillation, which is another key factor associated with OHCA survival. 28 The bystander defibrillation rate worldwide was 2%–9% of all OHCA.29–32 The much lower utility of public access defibrillator may be explained by the inaccessibility, difficulty in locating AEDs and a lack of awareness among untrained bystanders of the need of AED use in resuscitation. Actions should be taken to improve the PAD rate, including strategies to optimise AED deployment in the community and mobile applications to shorten the time in locating the nearest AED. These measures should be bundled with the promotion campaigns on DA-CPR and PAD in the community.
On the other hand, from our study, most of the OHCA took place at home or elderly home, while only 11% occur in public place. Apart from promoting PAD use, another question to answer is how to improve AED accessibility for OHCA in residential place. Study by Jørgensen et al. 33 in Vienna and Copenhagen in 2022 suggested that deploying AED in social housing may be a feasible strategy for optimising coverage of residential OHCA.
Our study is the first territory-wide study on OHCA survival in Hong Kong after implementation of the PDA system by the HKFSD, and it has provided some insights to the benefits and barriers of the system in our locality. Yet there were several limitations of the study. First, missing data rate is high (21.9%). The main reason for the missing data is the mismatch between the ambulance record number and relevant hospital identity number resulting in failure in retrieving data from the CMS. Moreover, due to the retrospective design of the study, there were some incomplete data items which the investigators were unable to trace the data despite efforts. For instance, the caller characteristics were determined largely based on the conversation in the dispatch recordings. The relationship between callers and OHCA victims and callers’ CPR training background may affect the provision of bystander CPR, but we could not ascertain these factors in our study.
Second, the time from cardiac arrest to emergency call might not be accurate as it was estimated by the callers who might not have received relevant training and many arrests were not witnessed. Third, we could not assess the quality of bystander CPR, which is critical in OHCA survival. Fourth, there was lack of prior sample size calculation. Instead, we selected a 2-month period for analysis. Our study would be regarded as a pilot local study to investigate the effect of PDA on OHCA survival at our locality. It will provide insight for future study. The short duration and small sample size limited the impact of our study. According to our result, presuming the effect size remains unchanged, the calculated sample size shall be 5228, given alpha cut-off of 5% (0.05) and beta cut-off of 20% (0.2).
The lack of significant difference in the ROSC and survival rates between treatment and control groups may be attributed to the small sample size. We did not evaluate STD with a favourable neurological outcome because the number was too small for any meaningful comparison.
Finally, we did not evaluate the treatment after ED arrival, which might also affect the survival outcome. Despite these limitations, our study provides an update of the OHCA situation after implementation of PDA and empirical evidence to support DA-CPR in improving the prehospital chain of OHCA survival in Hong Kong. Future study should target at a longer study duration in a prospective design and adequate sample size to demonstrate the impact of DA-CPR on improving the survival of OHCA. Also in view of the relative old age of the population, it may generate more relevant results if data from different age groups are collected and compared.
Conclusion
In Hong Kong, PDA improved the overall bystander CPR rate in OHCA and shortened the time from cardiac arrest recognition to chest compression, but the improvement in survival did not reach statistical significance. More public education is necessary to provide information about how to recognise cardiac arrest and what occurs in DA-CPR, and to improve CPR and PAD knowledge and skills in the community while addressing the barriers.
Footnotes
Author contributions
Declaration of conflicting interests
Funding
Availability of data and materials
Ethical approval
ORCID iDs
