The project will use large scale survey with multiple waves of self-reported data, linked to routinely collected datasets. Multiple imputation algorithms will be developed and evaluated for use in epidemiological studies.