If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
1 This author takes responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation
C. Raina MacIntyre
Footnotes
1 This author takes responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation
Affiliations
Biosecurity Research Program, The Kirby Institute, UNSW, Sydney, AustraliaSchool of Public Affairs, College of Public Service and Community Solutions, Arizona State University, USA
The peak cardiac arrest hospitalisations overlapped with the seasonal peak of influenza cases.
•
The association between influenza and sudden cardiac arrest (SCA) is highest in adults 65 years and older.
•
Understanding the cardiac burden attributable to influenza could inform the use of influenza vaccine to reduce SCA in adults.
Abstract
Background
Cardiac arrest is the least preventable burden of cardiovascular disease, as treatment depends on timely resuscitation. The incidence of sudden cardiac arrest (SCA) is high, contributing 10–20% of cardiovascular mortality globally. The influenza vaccine reduces the risk of acute cardiovascular events. Little is known about the relationship of influenza infection to cardiac arrest.
Methods
This study aimed to determine the estimated rate of SCA hospitalisations attributable to influenza in Australian adults. A generalised-additive statistical model was applied in the study. Weekly counts of laboratory-confirmed influenza notifications were used as independent variables in the model.
Results
Our estimates showed that the yearly rate of SCA hospitalisations varied, and a significant association with influenza was observed in some years in older adults aged 65 years and over. On average, the annual estimated SCA hospitalisations rate due to influenza in adults aged 50–64 years and ≥ 65 years were 0.7 (95%CI: 0.4, 1.1) and 5.3 (95%CI: 4.4, 6.2) per 100,000 population, respectively.
Conclusion
The association between influenza and SCA is evident in adults and the disease burden is significant in older people. Prevention of influenza by vaccination may reduce SCA.
]. There are improvements in cardiopulmonary resuscitation (CPR) awareness and post-resuscitation care in recent time, however survival rates following SCA remain low, with an estimated 7.1% of patients surviving until discharge globally [
]. This is because survival depends on resuscitation within six minutes of SCA (the time to irreversible brain damage), and many people are not resuscitated on time.
There is evidence that influenza can trigger acute cardiovascular events such as ischaemic heart disease [
]. Clinically, it can manifest a range of acute coronary syndromes including angina, acute myocardial infarction (AMI) or malignant arrhythmias resulting in cardiac arrest [
]. Over time, both the incidence and mortality of AMI have decreased steadily across Australia due to improvement in prevention and management strategies and rapid revascularisation [
The global survival rate among adult out-of-hospital cardiac arrest patients who received cardiopulmonary resuscitation: a systematic review and meta-analysis.
]. Eighty percent of SCA have a cardiac cause, with majority of these cases being due to coronary artery disease, and influenza may affect SCA rates by triggering AMI [
]. A study of 163,831 patients with implantable cardiac defibrillators concluded that high influenza rates were associated with a high rate of ventricular arrhythmias requiring treatment with shock or anti-tachycardia pacing [
Effect of high influenza activity on risk of ventricular arrhythmias requiring therapy in patients with implantable cardiac defibrillators and cardiac resynchronization therapy defibrillators.
]. Whilst the association between influenza and arrhythmias is still poorly understood, there is evidence that higher influenza activity may be linked to high incidence of supraventricular tachycardia episodes [
]. A time-series analysis of 481,516 cases of out-of-hospital cardiac arrests has further suggested a link between severe influenza epidemics and an increase in cardiac arrests [
]. However, despite the high mortality and morbidity rates associated with the condition, there is limited data on the relationship between influenza and SCA.
Increases in influenza incidence are associated with a 2.3–6.3% increase in cardiovascular disease mortality in patients aged 65 years and over [
]. Understanding the disease burden of SCAs related to influenza will better inform potential prevention with influenza vaccines. We aimed to estimate the rate of cardiac arrest hospitalisations attributable to influenza in adults aged 50–64 years and ≥ 65 years in Australia from 2001 through 2017.
2. Methods
2.1 Data source
National hospital admissions data were extracted from the Australian Institute of Health and Welfare (AIHW), Australia. Weekly counts of ICD-10 AM coded hospital admissions with a principal diagnosis of cardiac arrest (I46) were extracted under two age groups: 50–64 years and ≥ 65 years between 2001/02 and 2017/18 in the study [
]. These age groups were selected in the study because adults aged 65 years and over are eligible for free influenza vaccine under the national program, whereas those under 65 years are only eligible for free vaccination if they have other risk factors or comorbidities [
]. Also, weekly counts of all-age laboratory-confirmed influenza notifications between 2001 and 2018 were collected from the National Notifiable Diseases Surveillance System (NNDSS) [
]. Influenza is a notifiable disease and all cases are notified by authorities across all states and territories to the national surveillance system in Australia. The population data was sourced from the Australian Bureau of Statistics (ABS) to calculate the age-specific rate [
Descriptive epidemiology and statistical modelling were used to estimate the rate of cardiac arrest admissions attributable to influenza in adults 50–64 and ≥ 65 years of age. A semi-parametric generalised-additive model (GAM) with Gaussian distribution was applied in the study [
]. Yearly influenza notification data was used as an input variable in the model. The outcome variable for the time series model was weekly rates of cardiac arrest hospitalisations, and we estimated the age-specific rate of influenza-attributable cardiac arrest hospitalisations for each age group separately by the model. An interaction term between year and influenza notifications was used in the model by putting notifications as a separate variable for each year. It has been used to adjust for variation in reporting practices across all states and territories over time. We analysed data in two parts: pre-pandemic (the period prior to 2009, July 2001–December 2008) and post-pandemic models (the period including January 2009–June 2018). It was done due to significant changes in notification trends between the two periods [
]. Further, a natural cubic smoothing spline of the consecutive week number was included in the model to account for unmeasured non-influenza-attributable cardiac arrest hospitalisations due to other unknown factors. The following equation was used in the model:
where Y represents the estimated weekly cardiac arrest hospitalisation rate per 100,000 population for a specific age group. β0 is the model intercept, β1, year represents yearly notifications, β2 (time) is a linear term for week number to control for any long-term (ecological) linear time trend in the hospitalisation rate, and spline (time) is a smoothing spline of the week number to control for unmeasured time varying factors such as non-influenza related seasonal variation in the hospitalisations. Each time series had 887 observations covering all weeks continuously during the study period. For smoothing spline, shorter intervals with flexible mode, i.e., six knots per year was applied in the model. A total of 46 and 58 degrees of freedom including one degree of freedom for a linear time was specified for two study periods (pre- and post-pandemic years), respectively. This choice was done based on published studies previously [
Weekly hospitalisation rates were summarised to yearly rates and an annual average rate per 100,000 population was calculated for each age group excluding the 2009 pandemic year in the study. For statistical significance, 95% conference interval (CI) was applied (included the values that were not overlapping zero), using the formula, β ± (1.96 x SE), where “SE” is the standard error of the parameter estimate from the model.
An average CI was calculated as per formula below, by using the square root of the sum of the squared standard errors of parameter estimates for each annual influenza variable, divided by the number of years included in the average exclude 2009 [
where SE1 … SEn are the standard errors of the parameter estimates for the influenza variables in each of the n years averaged.
Normal distribution of data was assumed, and quantile-quantile (QQ) plot was checked for normality of data in the model residuals. Autocorrelation of time series data was checked using the autocorrelation plots. For analysis, the SAS Enterprise Guide statistical software (SASEG) version 8.1 (SAS Institute Inc., Cary, NC, USA) was applied in the study.
3. Results
During the study period, the total number of hospitalisations with principal diagnosis of cardiac arrest was 30,822 for 50–64 years and 91,205 for adults aged ≥65 years. Average annual rates of cardiac arrest hospital admissions were 47.6 and 185.6 per 100,000 population in adults 50–64 and 65 years and over, respectively. There were variations in rates of cardiac arrest hospitalisations over time in both age groups. From the observed data as shown by boxplots in Fig. 1, weekly trend of cardiac arrest hospitalisations in adults aged 65 years and over revealed a seasonal pattern (Fig. 1A), however, no apparent seasonality was seen in adults 50–64 years in the study (Fig. 1B). For the model fit, autocorrelation plots showed statistically significant mild correlations in the model residuals. The QQ plots presented a better model to the observed data with some departures at the extreme values in both age groups.
Fig. 1Observed weekly hospitalisation rate per 100,000 population with principal diagnosis of cardiac arrest, 2001–2017 highlighting a seasonal trend in cardiac arrest hospitalisations in adults 65 years and above (A) but not in those 50 to 64 years (B).
The reported number of all age laboratory-confirmed influenza notification rates are presented in Fig. S1 (supplementary). The graph shows seasonal trend of influenza infection by month and year, where low to moderate or high influenza activity was reported during these seasons in Australia. The notification showed significant changes in pre- and post-pandemic years, as shown in Fig. S1 a / b. The seasonal peak was seen around August–September, except 2009. Across the years, the highest peak was seen in 2017 with a notification rate of 404.3 per 100,000 population, followed by rate of 164.6, 121.7 and 114.2 per 100,000 population in 2015, 2016 and 2014, respectively. On average, a seasonal peak notification rate of 63.3 per 100,000 population (excluding the 2009 pandemic year) was reported in August during the study period.
3.1 The estimated annual hospitalisation rate
The estimated cardiac arrest hospitalisation rate attributable to influenza varied by year and age group and is presented in Table 1. During the pre-pandemic period, rates of influenza attributable cardiac-arrest admissions in adults aged ≥65 years were high: 15.6 (95%CI: 12.0, 19.2) in 2002 and 14.3 (95%CI: 10.3, 18.2) in 2005 per 100,000 population, respectively. In post-pandemic years, rates of 6.4 (95%CI: 3.1, 9.7) in 2012, and 4.6 (95%CI: 1.4, 7.7) in 2014 per 100,000 population were seen among the older adults. In adults aged 50–64 years, no significant association was found in the study except for some years, as shown in Table 1. In that age group, the highest estimated rate of cardiac arrest hospitalisation attributable to influenza was 2.0 (95%CI: 0.3, 3.7) in 2013, followed by a rate of 1.9 (95%CI: 0.6, 3.3) per 100,000 population in 2017.
Fig. 2 presents the estimated weekly rates of influenza-attributable cardiac arrest hospitalisations in two age groups in Australia compared to the baseline and the observed rate of cardiac arrest hospitalisations from 2001 through 2017. There were reductions in cardiac arrest hospitalisations in later years in both age groups in the study.
Fig. 2Estimated, observed and baseline influenza-attributable cardiac arrest hospitalisation rate* by age group and year, Australia, 2001–2017.
Note: Blue-coloured area indicates the rate of influenza-attributable hospitalisation, and a light grey area denotes the baseline hospitalisation rate per year in the graph.
* Rate per 100,000 population. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
3.2 The estimated average annual hospitalisation rate
On average, the estimated annual rate of cardiac arrest hospitalisation attributable to influenza was 5.3 (95%CI: 4.4, 6.2) per 100,000 population in adults aged ≥65 years. The corresponding rate of 0.7 (95%CI: 0.4, 1.1) per 100,000 population was found in adults 50–64 years during the study period. Older adults ≥65 years have a higher disease burden in general than adults aged 50–64 years across the studied years.
Fig. 3 shows the estimated monthly rate of cardiac arrest hospitalisations attributable to influenza, with corresponding CIs in two age groups by selected years. There was variation in monthly peak rate across the years. In the 50–64-year age group, the highest monthly seasonal peak was seen in 2017, with a rate of 0.73 per 100,000 population in September, followed by a rate of 0.70 per 100,000 population in 2003 in August (Fig. 3A). In adults aged ≥65 years, the highest seasonal peak was seen in 2002, at a rate of 5.7 per 100,000 population, while in 2017, a late peak was seen in September at a rate of 0.98 per 100,000 population (Fig. 3B).
Fig. 3Estimated influenza-attributable cardiac arrest hospitalisation rate by age group and by month, year*.
This study demonstrated the estimated burden of influenza-attributable cardiac arrest hospitalisations in Australia between 2001 and 2017. The annual rate of influenza-attributable cardiac arrest hospitalisations varies by age group. In older adults ≥65 years, it reveals a seasonal trend compared to 50–64 years old in the study.
The mechanism between cardiac arrest and influenza infection may not be known clearly; however studies demonstrated that influenza infection triggers cardiac events such as acute coronary syndrome, myocarditis, myocardial infarction, cardiac arrhythmias and sudden cardiac death [
]. Sudden cardiac deaths occur due to onset of fatal ventricular arrhythmias in those with underlying pathology of coronary heart disease, and are common in older adults but lower incidence (<1%) is reported among young adults aged <35 years [
]. A significant association between influenza and ventricular arrhythmias was reported in a modelling study that recruited over 160,000 adults during the influenza seasons from 2009 to 2015. In that study, authors found that incidence of ventricular arrhythmias cases was high during the seasonal peak [
Effect of high influenza activity on risk of ventricular arrhythmias requiring therapy in patients with implantable cardiac defibrillators and cardiac resynchronization therapy defibrillators.
One Australian study conducted in the State of Victoria from 2008 to 2017 demonstrated a higher incidence of sudden cardiac arrest cases during winters compared to summer periods (22 versus 18 per 100,000 population) [
]. Seasonal variation in incidence of OHCA has been reported in another study by El Sibai et al. from emergency department records from the United States in 2014. Their findings showed that December and January had the highest records of OHCA presentations with lower survival rates in winters than summer periods [
]. An association between severe influenza season and reporting of increased rate of cardiac arrest cases has also been shown in a study by Onozuka et al. [
] Our findings show a similar outcome to that study; where the peak rate of influenza-attributable cardiac arrest hospitalisation follows the seasonal epidemic trend, with rate of highest peak falls between August and October during the studied years, except 2009.
Clinical presentations of influenza A and B may be indistinguishable. Although some findings are controversial, varied clinical outcomes by different influenza strains and severe clinical outcomes of influenza with A(H3N2) strain have been documented by studies [
Estimates of mortality attributable to influenza and RSV in the United States during 1997–2009 by influenza type or subtype, age, cause of death, and risk status.
Estimates of mortality attributable to influenza and RSV in the United States during 1997–2009 by influenza type or subtype, age, cause of death, and risk status.
]. By subtype, our study estimated an increased rate of influenza-attributable cardiac arrest hospitalisations during the A(H3N2) virus predominated seasons. However, in adults aged 50–64 years, increased rates of influenza-attributable cardiac arrest hospitalisations were seen in 2013 and 2017, where A(H3N2) and A(H1N1)pdm09 viruses were high in circulation in those years.
The incidence of cardiac arrest cases varied by region globally. Among them, a population rate of 95.7 per 100,000 in north America [
EuReCa ONE—27 Nations, ONE Europe, ONE registry: a prospective one month analysis of out-of-hospital cardiac arrest outcomes in 27 countries in Europe.
], while rate of out of hospital cardiac arrests in 2015, 105.5 per 100,000 population in New Zealand and 99.4 per 100,000 population (which equals to 24,373 cases) in Australia [
Regional variation in the characteristics, incidence and outcomes of out-of-hospital cardiac arrest in Australia and New Zealand: results from the Aus-ROC Epistry.
] have been identified. The reported average cardiac arrest hospitalisation rate in adults aged 50–64 and ≥ 65 years in Australia were 47.6 and 185.6 per 100,000 population, and our study demonstrated that influenza was associated with 1.5% and 2.9% of total cardiac arrest hospitalisations in those age groups during the studied years, respectively.
The association between influenza and AMI has been widely studied [
], and AMI is the likely precursor and common attribute of SCA. In a population study of SCA from India, authors recorded that about 35% and 22% of SCA deaths had prior myocardial infarction and left ventricular dysfunction [
]. The highest impact of mortality was also seen in people aged 50–70 years old in that study. An association between influenza and myocardial infarction was found in our previous modelling study; the estimated rate of influenza-attributable myocardial infarction for the same study period (2001–2017) in adults 50–64 years was 1.2 (95%CI: 0.1, 2.4) and a rate of 13.9 (95%CI: 11.2, 16.6) per 100,000 population was seen in older people ≥65 years [
]. A significant reduction in cardiovascular events is reported in a meta-analysis of randomised controlled trials, with reduced risk of cardiac events by 2.9% among participants who received influenza vaccine compared to 4.7% in the placebo group [
]. Another meta-analysis on influenza vaccine and cardiovascular events demonstrated the protective effect of vaccine, and authors stated that influenza vaccine could reduce cardiovascular-related hospitalisations by 16%, cardiac-related mortality by an average of 56% and all-cause mortality by 22–40% in the study [
]. Given the low survival rate of SCA, influenza vaccine could potentially prevent death from SCA.
In Australia, annual vaccination is recommended for all people ≥6 months old for infection prevention; however, low vaccine coverage was reported among adults in general [
]. A 2014 vaccination survey of adults revealed that only 39% of all Australian adults had an influenza vaccination in 2014, and vaccine coverage among 50–64 years in that same year was 46% compared to 73% in adults aged ≥65 years. The reported vaccine coverage among the at-risk population in 2014 was 63% in the study [
]. Thus, low vaccine coverage in adults <65 years of age or the targeted high-risk population could result in additional healthcare and economic burden related to influenza.
5. Limitations
There are limitations in the study. Firstly, laboratory-confirmed notifications across the studied years varied over time, and testing and reporting practices varied nationally, affecting the model outputs. Secondly, the study model only considered acute implications of influenza infection, and long-term consequences of infection were not accounted for, hence, a limitation in the study. Thirdly, all-age influenza data was used as input variables in the model due to limitations or smaller amount of data available as age-specific influenza data in the study. We anticipated that age-specific influenza data could help in model estimates precision, however further research is necessary. Finally, other common seasonal respiratory infections, such as respiratory syncytial viruses were not adjusted for in the model, and as such our estimates may be overestimated, and study findings should be interpreted with caution.
6. Conclusion
In conclusion, our study indicated an association between influenza and cardiac arrest admissions in older adults aged ≥65 years. The peak incidence of cardiac arrest hospitalisations overlapped with seasonal peak across the years in the study. Understanding the cardiac burden attributable to influenza such as AMI, a common precursor of SCAs among the high-risk population, could inform the use of influenza vaccine to reduce SCA in adults.
Acknowledgement of grant support
This research did not receive any specific funding.
C. Raina MacIntyre is supported by a NHMRC Principal Research Fellowship, grant no. 1137582.
Declaration of Competing Interest
C. Raina MacIntyre reports grants from Sanofi and other support from Seqirus, AstraZeneca Australia, and Janssen, outside of the submitted work.
Acknowledgements
Hospitalisation data was supported by the Australian Institute of Health and Welfare, and influenza surveillance data was from the NNDSS, Department of Health, Australia.
The global survival rate among adult out-of-hospital cardiac arrest patients who received cardiopulmonary resuscitation: a systematic review and meta-analysis.
Effect of high influenza activity on risk of ventricular arrhythmias requiring therapy in patients with implantable cardiac defibrillators and cardiac resynchronization therapy defibrillators.
Effect of high influenza activity on risk of ventricular arrhythmias requiring therapy in patients with implantable cardiac defibrillators and cardiac resynchronization therapy defibrillators.
Estimates of mortality attributable to influenza and RSV in the United States during 1997–2009 by influenza type or subtype, age, cause of death, and risk status.
EuReCa ONE—27 Nations, ONE Europe, ONE registry: a prospective one month analysis of out-of-hospital cardiac arrest outcomes in 27 countries in Europe.
Regional variation in the characteristics, incidence and outcomes of out-of-hospital cardiac arrest in Australia and New Zealand: results from the Aus-ROC Epistry.
Rate per 100,000 population.
00631-3&id=gr1.jpg){kind=link}
00631-3&id=gr2.jpg){kind=link}
00631-3&id=gr3.jpg){kind=link}