Effects of atrial fibrillation on outcomes of influenza hospitalization.

Background: There is little information available on AF and its association with outcomes in adult influenza hospitalizations. Methods: The National Inpatient Sample was queried from years 2009-2018 to create a cohort of discharges containing an influenza diagnosis. AF was the primary exposure. Univariate and multivariate regression analysis was used to describe the association of AF with clinical and healthcare-resource outcomes. Finally, a doubly-robust analysis using average treatment effect on the treated (ATT) propensity score weighting was performed to verify the results of traditional regression analysis. Results: After adjustment, the presence of AF during influenza hospitalization was associated with higher odds of in-hospital mortality (aOR 1.56, 95 % CI 1.49 - 1.65), acute respiratory failure (aOR 1.22, 95 % CI 1.19 - 1.25), acute respiratory failure with mechanical ventilation (aOR 1.37, 95 % CI 1.32 - 1.41), acute kidney injury (aOR 1.09, 95 % CI 1.06 - 1.12), acute kidney injury requiring dialysis (aOR 1.61, 95 % CI 1.46 - 1.78) and cardiogenic shock (aOR 1.90, 95 % CI 1.65 - 2.20, all p-values < 0.0001). These findings were validated in our propensity score analysis using ATT weights. The presence of AF was also associated with higher total charges and costs of hospitalization, as well as a significantly longer length of stay (all p-values < 0.0001). Conclusion: AF is a cardiovascular comorbidity associated with worse clinical and healthcare resource outcomes in influenza requiring hospitalization. Its presence should be used to identify patients with influenza at risk of worse prognosis.