DESCRIPTION (provided by applicant): Because of near universal recommendation for annual vaccination (all aged 6 months) only observational studies of influenza vaccine effectiveness can be performed for ethical reasons. However, results from observational studies may be subject to unrecognized selection biases, and study design strategies to identify and adjust for these biases are necessary. We propose estimation of influenza vaccine effectiveness in preventing medically attended influenza in two health systems, and influenza of all severities in a cohort of households with children, in Michigan. We will conduct surveillance at multiple outpatient health centers, three hospitals, and in ~300 enrolled households during the influenza season, and will enroll adult and pediatric patients and household members meeting various case definitions. Vaccination status will be reported and validated, and considered with laboratory-confirmed influenza outcomes to estimate vaccine effectiveness separately for ambulatory, inpatient and household populations, with estimates compared across studies. Analyses will use a test-negative control approach-those testing positive for influenza are cases, those testing negative are controls. Modifiers and confounders of vaccine effectiveness such as age, health status, household size and composition, health care utilization, subjective socio-economic status, and health system characteristics will be assessed and considered in analytic models.

Relevance: MFIVE-HIVE combined Annual estimation of influenza vaccine effectiveness is necessary to guide vaccination and control policy. We propose estimation of influenza vaccine effectiveness in preventing medically attended influenza among patients in two health systems, and influenza of all severities in a cohort of households with children, in Michigan. Laboratory confirmed influenza outcomes and verified influenza vaccination status will be used to calculate vaccine effectiveness using a test- negative control approach.