RCTs Flashcards
Including randomisation, blinding, allocation concealment, bias
What is it called when the sequence of study participants in each study arm is with-held from investigators?
Allocation concealment.
What is the process called when outcome assessors and/or study subjects are not aware of the intervention groups study participants are in?
Blinding.
What is it called when analysing RCT data based on the original groups the participants are allocated to?
Intention to treat.
What is the process called that aims to create two similar study groups and hence control for known and unknown confounders?
Randomisation.
Randomised controlled trials can test what type of relationship between a study factor and an outcome factor?
Causal.
What is needed to ensure true randomisation?
Allocation concealment.
What can RCTs determine?
Causality Magnitude Type Duration Frequency
What forms of bias occur in RCTs?
- Selection bias
- Confounding
- Attrition bias (loss to follow up) - can compromise validity.
How do you deal with confounding in RCTs?
Randomisation.
How do you deal with selection bias in RCTs?
Allocation concealment and blinding.
How do you deal with attrition bias (loss to follow up) in RCTs?
Properly calculating the loss to follow-up can only be done by determining the right denominator (includes all those randomly assigned).
<5% loss leads to little bias, while >20% poses serious threats to validity.
