HLSC 4P99 Lecture Notes - Lecture 7: Adobe After Effects, Cardiovascular Disease, Perfusion
Document Summary
Trial and error (tamoxiphan prevents breast cancer in the other breast) Epidemiological evidence (fluoride and water, red wine a cvd) Disadvantage: so many outside variables that influence results (systematic bias) (type 1 error, alpha error) We allow only 5% error for alpha, and 15% error for beta error. Important subjects reflect patients who get the diseases. Distinguishes the difference between population epidemiology and clinical epidemiology. As long as we can justify it, it"s pretty cool. Example: thyroid surgery works 95% of the time, lumbar perfusion surgery <3% of time. Problematic when the participants know they are being watched. From a physician"s perspective this is problematic- we"re supposed to help the patient, but if it works, hey, awesome! Match base on the usual care, then give one group the new medication. Downside is the aftereffects that may come up later. Intended to identify a dose range that is well tolerated and safe (high frequency severe side effects)