What Is the Limitation of a Retrospective Cohort Study?

i would describe this as a negative trial, but don't know what you thought about it. They found that the patients who were given the shorter courses verse the longer courses, they did not have increased mortality. No matter how good of a job we tryd to do to control for confounders, there's always going to be this problem of residual confounding. i love the example because it kind of high lights how, even if you match on all these things that you can measure, there's still going to be some unmeasurable differences that you're not able to capture with variables like that.