Influenza virus infection may increase risk of acute myocardial infarction
In 2018, a health-database study by Kwong et al. reported a 6-fold increased risk of acute myocardial infarction (AMI) during the week following a confirmed influenzavirus infection.
The database did however not include mortality information, essential to account for censoring and capture out-of-hospital AMI deaths. The aim of this study is to replicate the study by Kwong et al., but improve analyses by including mortality data.
Study design and methodology were similar to Kwong et al., except for the addition of mortality information. Laboratory records on respiratory virus PCR-testing from 16 laboratories across the Netherlands were linked to national mortality, hospitalization, and administrative registries. Influenza virus infection was defined as a positive PCR-test. AMI was defined as either a hospitalization or death registered as ICD-10 code I21.
Influenza virus infections during the AMI hospitalization were excluded since no temporal relation could be established. Using a self-controlled case-series design, we compared AMI incidence during the risk interval (day 1-7 after influenza virus infection) to the control period (one year before and after the risk interval).
Results
Between 2008 and 2019, 401 patients were identified with AMI within one year before or after an influenza virus infection.
Of these, 139 (34.7%) patients died (all-causes) within one year after influenza virus infection.
25 AMI cases occurred in the risk interval versus 217 in the control period before and 177 after influenza virus infection.
The adjusted incidence ratio (IR) of AMI in the risk interval compared to control was 6.16
Excluding AMI cases detected from death records, similar to the original study, reduced the IR to 2.42
Using similar methodology as the original study in a different population, a significant but weaker association between influenza virus infection and AMI was found, confirming the generalizability of myocardial risk from influenzavirus infection.
By including mortality data, the IR increased more than two-fold, illustrating the impact of information bias on estimation.
Compared to the original study, the difference in results may be further explained by difference in testing practices and organization of care with limited influenza virus testing outside hospital setting.
References:
Kwong JC, Schwartz KL, Campitelli MA, Chung H, Crowcroft NS, Karnauchow T, Katz K, Ko DT, McGeer AJ, McNally D, Richardson DC, Rosella LC, Simor A, Smieja M, Zahariadis G, Gubbay JB. Acute Myocardial Infarction after Laboratory-Confirmed Influenza Infection. N Engl J Med. 2018. Jan 25;378(4):345-353. doi: 10.1056/NEJMoa1702090
