Safety and effectiveness of neuraminidase inhibitors in situations of pandemic and/or novel/variant influenza: a systematic review of the literature, 2009-15

Objectives: To review systematically the published literature evaluating neuraminidase inhibitor (NI) safety and effectiveness in situations of pandemic and novel/variant influenza. Methods: We searched six online databases using comprehensive search criteria for observational studies and randomized controlled trials investigating the effects of NI treatment, prophylaxis or outbreak control in patients of all ages. Results: Overall, 165 studies were included (95% observational), which were generally of low methodological quality due to lack of adjustment for confounding variables. In studies reporting adjusted estimates in general populations, NI treatment appeared likely to be effective against mortality (primarily if administered within 48 h of symptom onset) and potentially effective in reducing pneumonia. NIs appeared effective in reducing secondary transmission when indicated for prophylaxis. Limited, low-quality data suggest NIs are likely safe in general populations and may be safe in pregnant women and children. Data are scarce regarding safety of NIs in adults and high-risk individuals. Conclusions: Most included studies were observational, statistically underpowered and at high risk of reporting biased and/or confounded effect estimates. NI treatment appeared likely effective in reducing mortality (cause unspecified) and pneumonia in general populations, with increasing benefit when administered with 48 h of symptom onset. NI pre- or post-exposure prophylaxis is likely effective in reducing secondary transmission of influenza in a general population. Our evidence suggests NIs are likely safe to use in the general population; however, data for children and pregnant women are limited. Knowledge gaps persist in specific populations such as Aboriginals, high-risk individuals and the elderly.