Abstract
BACKGROUND: Current studies regarding glucocorticosteroid treatment of influenza have only estimated risk of critical illness or death which can be easily confounded by timing of treatment administration. We used severe acute respiratory infection (sARI) as an endpoint and investigated risk associated with receiving glucocorticosteroids before sARI onset. METHODS: sARI cases were defined as influenza-like illness (ILI) with pH1N1 infection and respiratory distress. Controls were defined as pH1N1 cases other than sARI and randomly selected from the community. We compared glucocorticosteroids and other medications used before sARI onset using a matched case control study adjusted for age group as well as underlying disease. Time-dependent risk and dose responses at different time periods over the course of sARI cases were also examined. RESULTS: Of the sARI cases, 34% received glucocorticosteroids before sARI onset compared to 3.8% of controls during equivalent days (OR(M-H) = 17,95%CI = 2.1-135). Receiving glucocorticosteroids before sARI onset increased risk of developing subsequent critical illness or death (OR(M-H) = 5.7,95%CI = 1.6-20.2), and the OR(M-H) increased from 5.7 to 8.5 for continued glucocorticosteroid use after sARI onset. However, only receiving glucocorticosteroids after sARI onset did not increase risk of severe illness (OR(M-H) = 1.1,95%CI = 0.3-4.6). Each increase in glucocorticosteroids dose of 1 mg/kg/day before sARI onset resulted in an increase of 0.62 (R(2) = 0.87) in the pMEWS score at the time of sARI onset. CONCLUSIONS: Early glucocorticosteroid treatment increased risk of sARI and subsequent critical illness or death; however, only receiving glucocorticosteroids after sARI onset did not increase risk of severe illness.