HTHSCI 3C04 Chapter Notes - Chapter 16: Sampling Bias, Coin Flipping
27 Mar 2018
School
Department
Course
Professor
Document Summary
Chapter 16 assessing allocation concealment and blinding in. Primary question for assessing validity of findings: whether assignment of patients to treatments was randomized and whether the randomization was concealed. Secondary question for assessing validity of findings: whether patients, clinicians, outcome assessors, and data analysts were unaware of patient allocation. Allocation concealment: only method to ensure groups being compared are equivalent at the beginning of the study, eliminates selection and confounding bias in rcts, success of randomization depends on 2 interrelated processes. Use a random process to ensure unpredictability of allocation sequence. E. g. computer-generated numbers, random number tables, coin flipping. Allocation concealment shields people in a trial from their assignments in advance. If randomization is based on whether patients" room numbers are odd or even, that information can be linked back to the patient and therefore is inadequate allocation concealment: strategies to conceal allocation. Calling a central coordinating office for patient assignment upon inclusion in the study.
