Non-specific benefit of seasonal influenza vaccine on respiratory syncytial virus-hospitalisations in children: An instrumental variable approach using population-based data

Seasonal influenza vaccine is effective against influenza hospitalisations, but little is known about non-specific effects of the vaccine on other respiratory pathogens with similar seasonal patterns. We aimed to assess the causal impact of seasonal influenza vaccine on laboratory-confirmed hospitalisations for respiratory syncytial virus (RSV) in children using an instrumental variable (IV) strategy.We used probabilistically linked population-based data on childhood immunisations, births, deaths, hospitalisations, perinatal factors, and microbiology test results (2000-2013) of all Western Australian (WA) children born 2000-2012, observed longitudinally until the earliest of 7 years of age or 31 December 2013. We exploited a unique natural experiment created from the WA's state-funded preschool influenza vaccination policy commencing in 2008 and used this as an instrument for children's seasonal influenza vaccination status. We estimated a system of two simultaneous probit equations: determinants of influenza vaccine uptake, and determinants of RSV-confirmed hospitalisation.Influenza vaccine coverage was low prior to 2008 but increased to 36 % in children aged 6-23 months in 2009. The majority (90 %) of RSV-hospitalisations occurred in children <2 years. Receipt of influenza vaccine reduced RSV-hospitalisations, especially in those <2 years with a rate reduction of 2.27 per 1000 (95 % CI: -3.26; -1.28), and a smaller rate reduction of 0.53 per 1000 (95 % CI: -1.04; -0.02) in those 2-7 years. Over the 5-year period (2008-2013), the state-funded preschool-influenza vaccine program resulted in 1,193 fewer RSV-hospitalisations. Of these, 793 (67 %) were in young children <2 years.To our knowledge, this is the first analysis utilising an IV estimation strategy on a population level to assess the causal impact of seasonal influenza vaccine on risk of RSV-hospitalisations. We estimated a small protective effect that warrants further investigation.