Return to Article Details Considerations of response bias and value in linking an existing longitudinal cohort study with national health record data Download Download PDF