Boudreault 2011.

Methods	Single‐centre retrospective cohort
Participants	Country: USA (Washington) Setting: in‐hospital Number of individuals: 143 Inclusion criteria: individuals undergoing haematopoietic stem cell transplantation presenting with respiratory tract infections Definition of influenza: laboratory confirmed Influenza type: seasonal Median age (years): cohort 42.0 (IQR 31.0 to 53.0); no CS group 42.0 (IQR 32.0 to 51.0); low‐dose CS group 42.0 (IQR 28.0 to 53.0); high‐dose CS group 40.0 (IQR 32.0 to 54.0) Male sex: cohort 83 (58.0); no CS group 36 (57.0); low‐dose CS group 29 (67.0); high‐dose CS group 18 (49.0)
Interventions	Groups: No CS (n = 63); low‐dose CS (n = 43); high‐dose CS (n = 37) Definitions for dose: low‐dose (prednisolone/methylprednisolone < 1 mg/kg/day); high‐dose (prednisolone/methylprednisolone >= 1 mg/kg/day) (dose considered was the highest dose taken during the 2 weeks preceding the influenza diagnosis) Co‐interventions: antiviral therapy
Outcomes	Time to death/time to influenza‐associated death: hazard ratios presented following multivariate analysis. Variables in the multivariate models included CS treatment, antiviral therapy, and lymphocyte count Hypoxaemia Lower respiratory tract disease Mechanical ventilation Adverse events: prolonged viral shedding
Risk of bias (Newcastle‐Ottawa Scale)	Time to death Selection domain score (max 4): 2 Comparability domain score (max 2): 1 (no adjustment for age/disease severity) Outcome domain score (max 3): 2
Notes	—