SAR HS 300 Lecture Notes - Lecture 5: Randomized Controlled Trial, Selection Bias
21 Mar 2019
School
Department
Course
Professor
Document Summary
Percentage of patients alive 5 years later: avg no. of years a patient survives after diagnosis. Perception that the screen-detected case has longer survival because the disease was identified early. Impact: can overestimate the benefits of a screening program when survival is used as the measure. Tumors identified by screening are slower growing and have a better prognosis. Impact: overestimate the benefits of the screening because screening tends to detect less aggressive forms of disease which have longer preclinical phase and better outcomes and longer survival. Selection bias: motivated participants who choose to participate may have a different probability of disease than do those who refuse to participate in screening. E. g. they could have better insurance or be more healthy. Despite intuitive appeal their utility and benefits should never be assumed. Screening tests must be evaluated with the same scientific rigor as any other aspects of epi investigation. Subjects receive either the new screening test or usual care.
