General practice encounters following seasonal influenza vaccination as a proxy measure of early-onset adverse events

• We investigated the utility of surveillance data as a means of signal detection.
• A unique identifier (Medicare number) linked GP consultation and ACIR datasets.
• A GP consultation on the day after influenza vaccine receipt is a reasonable proxy for early reactogenicity.
• This methodology could be used for signal detection of early-onset adverse events.

BackgroundIn 2010, use of seasonal trivalent influenza vaccine (TIV) in children <5 years of age was suspended in Australia following reports of vaccine-related febrile convulsions. We investigated the utility of data on primary care [general practice (GP)] consultations for any reason within three days of receipt of influenza vaccine as recorded on the Australian Childhood Immunisation Register (ACIR) as a means of signal detection.MethodsData on GP consultations were obtained from Medicare Australia (Australian Government Department of Human Services) for children recorded on the ACIR as receiving either TIV or monovalent influenza vaccine. Rates of GP consultation by day following ACIR-recorded receipt of influenza vaccine were compared by year (2008–2010), vaccine type, age and region.ResultsIn 2010, GP encounter rates on the day after receipt of the TIV manufactured by bioCSL (formerly CSL Biotherapies (Fluvax®) were significantly higher than both bioCSL TIVs in the previous two years [rate ratio (RR) 1.9; 95% CI: 1.7–2.2] and Sanofi Pasteur TIV, Vaxigrip® [RR 1.6, 95% CI 1.4–1.7] in 2009–2010. Encounter rates were also higher than for CSL Monovalent influenza vaccine, Panvax® [RR 1.9, 95% CI 1.7–2.2] in 2009–2010. These findings were robust to adjustment for age group (≤2, >2 years) and region (Western Australia vs other Australian states/territories).ConclusionsA primary care consultation on the day after vaccine receipt is a reasonable proxy for early reactogenicity and has potential for use in various settings.