Introduction:
In temperate regions, cardiovascular deaths and influenza epidemics peak with regularity during the winter months.
Hypothesis:
We assessed the hypothesis that population increases in seasonal influenza infections are associated with a rise in mortality due to cardiovascular causes, and that influenza incidence can be used to predict cardiovascular mortality rates during the influenza season.
Methods:
We used time series regression models, adjusted for season and time trend, to quantify the temporal association between influenza incidence and cardiovascular mortality during the influenza season in New York City. Mortality data on date of death, age, and underlying cause of death were obtained from the New York City Office of Vital Statistics. Daily mortality counts from 2006 to 2012 were aggregated for all cardiovascular causes (International Classification of Diseases, Revision 10 (ICD-10) codes I00-I99), ischemic heart disease (ICD-10 codes I20-I25), and myocardial infarction (ICD-10 code I21). Influenza incidence was represented using four different measures: emergency department visits for influenza-like illness, grouped by age ≥ 0 and age ≥ 65 years, and these same measures scaled by laboratory surveillance data for viral types/sub-types. The 2009 H1N1 pandemic period was excluded from temporal analyses and reserved for out-of-sample prediction.
Results:
There were 73,384 cardiovascular deaths among adults age ≥ 65 years during the influenza seasons between 2006 and 2012, excluding the 2009 H1N1 pandemic period. Interquartile range increases of the four indicators of influenza incidence in the previous 21 days were associated with increases in cardiovascular mortality of between 2.3% (95% confidence interval (CI): 0.7%, 3.9%) and 6.3% (95% CI: 3.6%, 8.9%), and increases in ischemic heart disease mortality of between 2.4% (95% CI: 1.1%, 3.7%) and 7.0% (95% CI: 4.1%, 10.0%). Associations were most acute and strongest for myocardial infarction mortality, with interquartile range increases for the four influenza indicators during the previous 14 days associated with mortality increases between 5.9% (95% CI: 2.7%, 9.2%) and 12.8% (95% CI: 5.1%, 20.6%). Out-of-sample prediction of cardiovascular mortality among adults age ≥ 65 years during the 2009-2010 influenza season yielded average estimates with 94.4% accuracy.
Conclusions:
Emergency department visits for influenza-like illness are associated with and predictive of cardiovascular disease mortality in New York City. Retrospective estimation of influenza-attributable cardiovascular mortality burden, combined with accurate and reliable influenza forecasts, could predict the timing and burden of seasonal increases in cardiovascular mortality.
Influenza-like illness-related emergency department visits: Christmas and New Year holiday peaks and relationships with laboratory-confirmed respiratory virus detections, Edmonton, Alberta, 2004-2014
