Randomised Control Trials Flashcards
What makes an RCT “controlled”
A comparison between the intervention we’re interested in and the one we hope it’ll better than
What 3 good reasons are there to randomise a controlled trial
- to eliminate bia in the allocation of treatments by removing selection and confounding biases
- to allow for blinding of patients, investigators, operators and outcome assessors
- to allow the use of probability theory to express the likelihood that a difference in treatments is due to chance only.
Why do we do patient blinding
to remove the potential placebo effect, if they knew they were in the experimental then they might get better just cus they think they should
Why do we do operator blinding where possible
So the operator doesn’t behave and treat the patient differently based on which technique they’re using
Why would be try to do researcher blinding
Something they’ve put a lot of time into so they want good results and may be tempted to alter allocation of patients and to exclude some from the analysis
Why would we try to do outcome assessor blinding
May unconsciously alter the way you test the outcome depending on what treatment was used, this blinding is considered one of the most important
What are the explanations for outcome differences when there is minimal risk of bias or confounding affecting the results
Casual - one treatment causes an improved result
Chance - there is no real difference between the treatments in their effectiveness but, purely by chance, the results suggest there is
What is probability theory
This is done by researchers to work out what the probability is that the results they’ve obtained are due to chance alone rather than due to there being a real difference between the 2 interventions
What does a p value of 0.01 represent
a 1% chance of something being due to chance
What p value is said to be statistically significant
less than or equal to 0.05, so less than or equal to 5% chance that the differences were due to chance
What is clinical significance
Its about the actual effect of the results, like plaque removal, defining what the actual effect was and the difference in the actual effect between different interventions
What is the ARR and RRR and how are they calculated
ARR = abolsute risk reduction, which is risk of one thing in one interventions subtracted from another
RRR = relative risk reduction, which is ARR divided by the risk of the control
Why is RRR preferred in abstracts and papers etc
the number looks more impressive very often
