HLSC 2P27 Lecture Notes - Lecture 8: Relative Risk, Logistic Regression, Statistical Hypothesis Testing
Document Summary
Collection of numbers we want to summarize for purposes of comparing that with numbers that are related. Need the five number summary of data. Last value carried forward: if we only have 9 months out of 12, and the last measurement was 100 we carry that value forward for the remaining 3 months. Best to avoid bias- cants inflate efficacy and avoid adverse events. Non-inferiority trials, both intention to treat and pre protocol analyses are recommended. How much is related to the exposure. P=0. 001, we can infer that the info is significant. Gives a difference in relative risk in two different studies. Covariates: random variables that affect each other, falls into the category of a confounding variable. Confounding variable: in rct we look for covariates that may be operate in the population overall, as confounders are strictly revised in the rct. Multivariable analysis: >= 2 independent (response) variables (variables that are effecting our outcome)