POPM 3240 Lecture Notes - Lecture 22: Aspirin, Pancreatic Cancer, Reye Syndrome
Document Summary
Some studies are not ethical to do for experimental studies. Prone to bias, less control than exp studies. Selecting subjects based on disease or outcome status. Cases are those with the disease, controls are those without the disease. Investigate whether the e is associated to o (whether the e differs between cases and controls) So therefore it is purposive sampling (can be random if a list exist but usually the diseased individuals are very few in number so we include every possible d+) Compare proportion of individuals e+ and e- by disease status. Retrospective, so it is fast and cheap. Best for outbreaks since a lot of people suddenly got sick and there are already a lot of people diseased. Good for rare diseases since based on disease status. Good for multiple risk factors (useful for disease of unknown etiology) Cannot calculate incidence or prevalence because we chose the prevalence when we chose number of cases.