Background
Influenza ranges from a mild and self-limiting infection to a life-threatening disease with high mortality despite intensive care. Conclusive data on the association between influenza type/subtype and mortality among adults treated at intensive care units (ICU) is lacking.

Objectives
To investigate the mortality in adults admitted to ICU with laboratory-confirmed influenza during three consecutive influenza seasons.

Methods
This observational multicenter study included adults with PCR-confirmed influenza requiring intensive care at four hospitals in southern Sweden between 2015–2018. The primary outcome was all-cause one-year mortality. Patient characteristics and the impact of influenza type/subtype were studied using Kaplan–Meier and logistic regression analyses.

Results
A total of 146 individuals were included: median age 67?years (interquartile range 56–74), 54% were male. Influenza type/subtype was available for 144/146 (99%); A(H1N1)pdm09 in 50 (35%), A(H3N2) in 37 (26%), and B in 57 (40%) patients. Mortality was 19% in the ICU and 32% before hospital discharge. At one year, 43% were diseased, ranging from 36% to 49%, depending on type/subtype (log-rank test p?=?0.32). Mortality rates remained similar for all three influenza types/subtypes after adjusting for age, sex, and a modified comorbidity index. Antibiotics were prescribed for 125/145 (86%) within 48?h of ICU admission, with microbiological confirmation of coinfection in 53/125 (42%).

Conclusions
Among adults admitted to intensive care with PCR-confirmed influenza, mortality rates were similar independently of influenza type/subtype. Mortality increased from 19% in the ICU to 43% one year after admission, highlighting the importance of monitoring ICU-survivors and reporting long-term outcomes in critically ill influenza patients.