Return to Article Details Potential pitfalls in observational study designs