STEPS IN PLANNING A COHORT STUDY
When to use a cohort design
- Accurately describing the incidence and natural history of a condition
- may be the only way to establish the temporal sequence of predictor and outcome variables.
- low conc of vitamin A and E in the serum of cases with cancer could be an effect of the cancer rather than its cause.
- only way to study certain rapidly fatal diseases.
- Avoiding the survivor bias
- allow investigators to study numerous outcome variables
- gains power to study an ever-increasing number of health outcomes.