Higher vaccine effectiveness in seasons with predominant circulation of seasonal influenza A(H1N1) than in A(H3N2) seasons: Test-negative case-control studies using surveillance data, Spain, 2003-2011

Background: We used data provided by the Spanish influenza surveillance system to measure seasonal influenza vaccine effectiveness (VE) against medically attended cases, laboratory confirmed with the predominately circulating influenza virus over eight seasons (2003-2011). Methods: Using the test-negative case-control design, we compared the vaccination status of swabbed influenza-like illnesses (ILI) patients who were laboratory confirmed with predominantly circulating influenza strain in the season (cases) to that of ILI patients testing negative for any influenza (controls). Data on age, sex, vaccination status and laboratory results were available for all seasons. We used logistic regression to calculate adjusted influenza VE for age, week of swabbing, Spanish region and season. We calculated the influenza VE by each season and pooling the seasons with the same predominant type/subtype. Results: Overall influenza VE against infection with A(H3N2) subtype (four seasons) was 31 (95% confidence interval (CI):10; 48). For seasonal influenza A(H1N1) (two seasons), the effectiveness was 86% (95% CI: 65; 94). Against B infection (three seasons), influenza VE was 47% (95% CI: 27; 62). Conclusions: The Spanish influenza surveillance system allowed estimating influenza VE in the studied seasons for the predominant strain. Strengthening the influenza surveillance will result in more precise VE estimates for decision making. (C) 2014 Elsevier Ltd. All rights reserved.