lecture 3 - randomised controlled trials Flashcards
What is clinical equipoise?
An ethical principle of randomised controlled trials that states that there must be genuine uncertainty about which intervention/treatment is better
What is a superiority trial?
A trial to determine whether a new treatment is superior to an existing treatment (or no treatment)
What is a non-inferiority trial?
A trial to evaluate whether a new treatment is no worse than an existing treatment
What is the limitation of using biomarkers as measures of outcome in RCTs?
They may not give an indication of clinical efficacy - e.g. quality of life, mortality
What is the key way that an RCT eliminates confounding?
By randomisation - divides patients into groups with the same distribution of characteristics
How can bias occur with the randomisation in an RCT?
If allocation is not blinded to researchers or patients, they may change their behaviour according to the group they are allocated.
How is allocation concealed in RCTs?
Blinding (of patient and researchers), placebo
How can bias be minimised during an RCT?
ensuring adherence to protocol (including making sure they have the assigned control or intervention), insuring retention/follow-up
What is intention to treat analysis?
If people don’t comply with their initially randomised groups, they are left in their original group for analysis
What is the purpose of intention to treat analysis?
Preserves randomisation and allows for better prediction of how treatments might work in a real life clinical setting, where there is unlikely to be complete adherence
What is a confidence interval?
A range of plausible values for the true effect of treatment in the target population
What is a p-value in an RCT?
A measure of the probability that we would have seen an effect as large or larger if the treatment did not actually work (i.e. the probability of a false positive result)
What value measures the probability of a falsely positive finding in an RCT?
p-value
What is the standard significance level of the p-value in an RCT?
p<0.05 = sufficient evidence of a treatment effect
What are the problems with multiple statistical analyses?
Inflation of chance of false positive test, interpretation problems with unplanned subgroups, bias due to subjective reporting of analyses favourable to treatment
