Abstract
Influenza remains a clinically significant viral cause of community-acquired pneumonia in adults. Without timely antiviral treatment, severe influenza complicated by pneumonia may lead to poorer outcomes. However, the effect of empiric antiviral treatment on serious or life-threatening influenza is not well documented from clinical trials. We performed a retrospective study of patients with severe influenza complicated by pneumonia as confirmed by bronchoalveolar lavage polymerase chain reaction (PCR) in intensive care units at a tertiary hospital in Taiwan from 2009 to 2019. We collected demographic and clinical data from medical records. We compared survival outcomes between those with or without empiric antiviral treatment (before or after PCR diagnosis) with the Cox proportional hazard model. Of the 77 patients enrolled in this study, the survival rates were higher among those treated empirically with antivirals before diagnostic test results were obtained than for those who were not (P = 0.024). Compared with the nonempiric treatment group, the empiric antiviral treatment group had a 68.3% lower risk of death. Meanwhile, the risk of death increased by 3.8% for each unit increase of acute physiology and chronic health evaluation (APACHE) III score. Empiric antiviral treatment of influenza may lead to a better survival rate. For patients with suspected severe influenza, antiviral medication should be initiated as soon as possible without waiting for laboratory confirmation of influenza virus infection.