Objective: To estimate influenza vaccine coverage and effectiveness against medically attended laboratory-confirmed influenza for the 2012 season.
Design, setting and participants: Test-negative design involving patients recruited as part of the Australian Sentinel Practices Research Network, a network of sentinel general practitioners throughout Australia. Throughout 2012, at the discretion of the GP at one of 102 participating practices, patients presenting with influenza-like illness were swabbed and included in the study.
Main outcome measure: Influenza vaccine effectiveness (VE) estimated as (1-OR)*100% by logistic regression.
Results: 1775 patients were swabbed. The epidemic period was identified as Weeks 10 to 43 of 2012. After exclusions, there were 1414 patients for the VE analysis, including 593 (42%) who tested influenza-positive and 821 who tested negative. 27% of test-negative patients were vaccinated, of whom most were aged 50 years and over. The overall VE, adjusted for age group, month of presentation and state or territory, was 23% (95% CI, -4% to 43%) against all influenza types, 15% (95% CI, -17% to 38%) against influenza A, 13% (95% CI, -20% to 36%) against influenza A(not H1) and 53% (95% CI, 5% to 77%) against influenza B.
Conclusion: Vaccination against influenza was modestly protective, reducing the risk of medical presentation with influenza by around 23%.