INTRODUCTION
Out-of-hospital cardiac arrest (OHCA) continues to impose a significant disease burden [
1,
2] worldwide [
3,
4] despite improvements in prehospital and in-hospital cardiac arrest responses over the past decades. The global survival rate of OHCA patients was found to be 7.7% at 1 year after hospital discharge [
5]. While there has been extensive research regarding short-term OHCA outcomes, relatively little is known about longer term outcomes in OHCA survivors. As increasing proportions of OHCA patients survive beyond 1 year in many communities around the world, there is emerging scientific interest in the long-term trajectories in the lives of OHCA survivors.
Strokes contribute significantly to the global burden of disease, being the second leading cause of deaths worldwide and the third leading cause of morbidity and mortality worldwide [
5]. In Singapore, cerebrovascular disease was the third leading cause of death for most years since 1970, comprising 9% to 12% of all deaths [
6]. Multiple studies have shown that increasing age is significantly associated with stroke prevalence. With Singapore’s aging population and high prevalence of stroke risk factors within the population, the incidence and health impact of strokes are expected to further increase. While number of studies look at long-term neurological outcomes and cognitive function in OHCA survivors, these mainly focus on memory and executive function as compared to the incidence of stroke.
This study aims to estimate the incidence of stroke in OHCA survivors in Singapore. We also looked to identify predictors or baseline characteristics of these patients that are associated with an increased risk of developing subsequent stroke.
METHODS
Study population
We obtained data from three different population-based registries: (1) Pan-Asian Resuscitation Outcomes Study (PAROS)–Singapore participants; (2) Singapore Stroke Registry; and (3) Singapore Registry of Births and Deaths. These were linked via the patients’ national registration identity card number, which is a unique identifier issued to every Singapore Citizen and Permanent Resident. Annual audits were carried out on data from each registry to ensure accuracy and reliability.
PAROS is an ongoing prospective multicenter registry of OHCA cases occurring in at least nine countries across the Asia-Pacific region [
7]. Only data from Singapore was used for the current study. All OHCA cases conveyed by emergency medical service or presenting at public emergency departments were included in this study. OHCA was confirmed by absence of pulse, unresponsiveness, and apnea. The PAROS data was then linked with both the stroke registry and the death registry to determine survival status and subsequent occurrence of stroke in the OHCA survivors.
The stroke registry is maintained by the National Registry of Diseases Office, Ministry of Health Singapore, and collects epidemiological data on Singapore residents (aged 15 and above) diagnosed with stroke [
8]. These cases are contributed by all public healthcare institutions in Singapore. Cases were identified using the International Classification of Disease 9th Revision (ICD-9) Clinical Modification codes 430 to 437 (excluding 432.1 and 435) prior to 2012, and the ICD-10 Australian Modification codes I60 to I68 (excluding I62.0 and I62.1) from 2012 onwards. All diagnoses of stroke were certified by a doctor based on clinical presentation. Dedicated registry coordinators extracted relevant clinical information after going through the patients’ electronic medical records. Further medical information was also obtained via ambulance, emergency department, and inpatient medical records or discharge summaries by registry coordinators.
The Singapore Registry of Births and Deaths is maintained by the Immigration and Checkpoint Authority and collects data on all live births, deaths and stillbirths occurring within Singapore and its territorial waters [
9]. This also traces the causes and dates of all deaths of Singapore Citizens and Permanent Residents, in accordance with the Registration of Births and Deaths Act 2021 [
10]. All identifiable data was removed by the Ministry of Health’s Unit for Prehospital Emergency Care prior to data analysis.
Study population and setting
Singapore is an urban city-state in Southeast Asia. Singapore has a unique healthcare system, whereby funding for public healthcare services is through a combination of compulsory savings, government subsidies, and price controls [
11,
12]. In 2020, cerebrovascular diseases (including stroke) were the fourth most common cause of death in Singapore [
13]. The two risk factors with the highest prevalence amongst stroke patients remained consistent from 2010 to 2020, with 82.9% of the patients with stroke in 2020 having hyperlipidemia and 81.5% having hypertension [
10].
Inclusion criteria and data collection
This study included all cases of OHCA in Singapore that occurred from April 1, 2010, to December 31, 2019, in the PAROS registry. From this population, patients who were not Singapore residents, younger than 18 years old, pronounced dead on scene or at the emergency department, or who died in hospital within 30 days from OHCA were excluded. We have provided a flowsheet showing selection criteria for OHCA cases in this study (
Supplementary Fig. 1).
Each episode of stroke was consistent with the World Health Organization’s MONICA (Monitoring Trends and Determinants in Cardiovascular Disease) criterion [
14]. Strokes in our defined population were ascertained and validated, whereby a recurring stroke within 28 days of a preceding episode was merged with the preceding episode, while a recurring stroke after 28 days of a preceding episode was counted as another stroke episode.
Statistical analyses
Baseline characteristics were presented as frequencies with proportions for categorical variables, and as median with interquartile range (IQR) for continuous variables. We used the Wilcoxon rank-sum test, Pearson chi-square test, and Fisher exact test to compare baseline characteristics between OHCA survivors with and without subsequent stroke where appropriate.
We performed Cox regression to identify factors associated with stroke after OHCA. The covariates investigated were: demographic factors (age, sex, and race), Utstein variables (bystander cardiopulmonary resuscitation [CPR], bystander automated external defibrillator [AED] use, CPR, bystander AED, witnessed arrest, first arrest rhythm, and prehospital return of spontaneous circulation [ROSC]), post-resuscitation therapeutics (such as emergency percutaneous coronary intervention, coronary artery bypass graft, targeted temperature management [TTM], or extracorporeal membrane oxygenation) and pre-existing comorbidities (previous acute myocardial infarction [AMI], previous heart disease, and previous stroke). For variables with few unknowns (n<10; bystander AED, prehospital ROSC), the unknown was excluded from the P-value calculation. For variables with substantial unknown (shockable first arrest rhythm, previous heart disease, and previous stroke), the unknown was included in P-value calculation. Since the event of death excludes a patient from a future stroke, competing risk from death was accounted by generating a sub-distribution hazard ratio for each factor in a Fine-Gray regression model. The level of significance was set at 5% and all analyses were performed using Stata V13 (Stata Corp.).
Ethics approval
This study was approved by the Centralized Institutional Review Board (CIRB Ref: 2018/2937) and was exempted from full review as only de-identified data was analyzed. The Ministry of Health’s Unit for Prehospital Emergency Care acted as a trusted third party to assist in removal of all identifiable information prior to data analysis.
DISCUSSION
In this nationwide, population-based study of OHCA patients who survived to 30 days post cardiac arrest, we found that: (1) 1.7% of OHCA survivors subsequently developed stroke; (2) previous stroke and AMI predisposes towards subsequent stroke after OHCA; or (3) TTM showed a trend towards decreased likelihood of subsequent stroke. While there has been increased interest in long-term outcomes of OHCA survivors in recent years, these mainly focus on neurological outcomes, cognitive function, and quality of life.
Current studies suggest a number of pathological links between stroke and cardiac arrest, with evidence pointing towards a higher incidence of cardiac death after an ischemic stroke. This is thought to be due to increased concentrations of cardiac enzymes and catecholamines, along with ongoing inflammation [
15]. Most of the existing studies focus more on cardiac events after stroke, and less so on the occurrence of stroke after an OHCA. To the best of our knowledge, this is the first and largest study specifically investigating the incidence and factors affecting subsequent development of stroke in OHCA survivors in a multi-ethnic Asian cohort [
16,
17].
Byrne et al. [
18] had a study population of 1,051 30-day OHCA survivors, and found that the absolute 5-year risk for OHCA survivors for stroke was 6.3% (95% CI, 4.1–8.5) as compared to 2.0% (95% CI, 1.6–2.5) for age- and gender-matched controls. Rey et al. [
19] assessed long-term clinical outcomes in OHCA survivors, with a total of 201 surviving patients included in the final analysis. They reported a 6.5% (13 patients) incidence of stroke at the end of follow-up, with a reported 93% of OHCA patients presenting with favorable neurological outcome throughout follow-up. These reported values are substantially higher than our reported stroke incidence of 1.7%. Reported crude incidence rate of stroke was 256 per 100,000 population in Singapore in 2021 [
8]. We postulate that the significantly different reported stroke incidence between our study and prior studies may be due to a number of different population and study factors.
A recent retrospective study in Singapore demonstrated that only 2.27% (260 out of 11,404) of OHCA survivors had favorable neurological outcome [
20], in comparison to Rey et al. [
19], where 93% of the study population had favorable neurological outcome. Although data on neurological outcomes was not included in our study, Liu et al.’s study [
20] was also a population-based study conducted in Singapore from April 2010 to December 2018 which largely corresponded with our study period and population. We postulate that the poor neurological outcomes in Singapore OHCA survivors likely contribute towards lower reported stroke rates due to difficulty in assessing for symptoms in an individual with pre-existing severe cerebral disability. Given the vastly different OHCA survivor to discharge rate of about 4% in Singapore as compared to 13.6% in Victoria or 11.7% in Europe [
21,
22], it is expected that many other system and population factors are unaccounted for. Further research is needed to better understand the factors affecting OHCA survival and stroke incidence in Singapore.
Increasing age significantly increased the risk of subsequent stroke in OHCA survivors (HR 1.05 [1.01–1.09]), which is consistent with current findings, with more than 70% of strokes occurring above 65 years old, and a doubling of stroke incidence with each decade after 45 years of age [
23]. Our findings of a trend towards increased incidence of stroke in patients with prior stroke (HR 2.20 [0.61–7.96]) and AMI (HR 2.96 [0.82–10.70]) are also congruent with prior studies demonstrating a strong association between AMI and stroke [
24]. Cerebrovascular disease and cardiovascular disease are both known to share many similar risk factors such as diabetes, hypercholesterolemia, hypertension and positive family history of heart disease [
15]. However further analysis into the trends of pre-existing comorbidities on stroke incidence was limited by the small number of patients with stroke post cardiac arrest.
TTM, also known as therapeutic hypothermia, involves the use of various cooling methods to induce hypothermia with the aim of improving neurologic outcomes. It is proposed that hypothermia reduces cerebral metabolic rate and oxygen demand, preserving energy stores, reducing lactate production [
25] and the further development of acidosis. The effectiveness of TTM has been demonstrated in multiple studies showing improved neurological outcomes and mortality in OHCA survivors [
26], and is also a part of recommended resuscitative guidelines for the International Liaison Committee on Resuscitation [
27] and the American Heart Association Emergency Cardiac care guidelines [
28]. In a randomized trial involving 77 comatose survivors of OHCA, treatment with moderate hypothermia was associated with improved outcomes [
29]. Similarly, TTM also demonstrated improved neurologic outcome (55% vs. 39%, P=0.009) and reduced mortality in patients with OHCA due to ventricular fibrillation [
30]. Implementation of TTM itself varies widely, with no current consensus as to the optimum temperature, duration, or specific cooling method. We found that OHCA survivors who underwent TTM showed a trend towards a lower risk of developing stroke (HR, 0.47; 95% CI, 0.10–2.12) (
Table 3). This is a novel finding demonstrating the benefits of TTM on the incidence of stroke post cardiac arrest, and further studies should be done to better elucidate the effect of TTM. As TTM was only implemented in 260 cases (29%) out of the 882 OHCA survivors, further improvements in protocol and processes may serve to further improve outcomes. More research and clinical trials would be essential to establish TTM as part of standard protocol.
To our best knowledge, this is the most extensive open cohort study summarizing the incidence, demographics and factors affecting subsequent development of stroke in OHCA survivors. Use of population-based registries also ensures that data obtained is accurate and standardized, and minimizes the likelihood of missing information or cases.
There are a number of limitations to this study. The small number of OHCA survivors that developed a subsequent stroke limited further analysis into the various patient and resuscitation factors in OHCA. Larger studies would allow for more accurate analysis and data interpretation to elucidate any underlying trends. Functional status of OHCA survivors are often assessed using cerebral performance category (CPC) or modified Rankin scale as per Utstein guidelines, which have been shown to be predictive of long-term survival [
31]. Information regarding CPC of OHCA survivors was not available in this study. This may have affected the accuracy of reported stroke incidence as patients with low CPC may not present with the stereotypical signs and symptoms of stroke, and subsequent deaths may have been attributed to other causes. As this was a retrospective cohort study, there may also have been confounding factors not accounted for. Other study methods such as the use of matched controls could also be considered in future. As our study was based on information collected from registries, information regarding patients who presented to private hospitals would not have been included.
In summary, incidence of stroke after OHCA was found to be 1.7% in our population-based cohort study in a multi-ethnic Southeast Asian country. OHCA survivors who developed subsequent stroke appeared likelier to have had a previous AMI or stroke, which is consistent with current studies demonstrating similar pathological risk factors toward the development of both cerebrovascular and cardiovascular disease. We also showed a trend towards decreased incidence of stroke after OHCA with use of TTM. Further studies on long-term outcomes of OHCA, in particular with relation to the incidence of stroke and TTM use may facilitate improvement in protocols and clinical care.