Randomised Control Trials Flashcards
What is an uncontrolled trial?
everyone gets the treatment
What is a controlled trial?
treated group compared with untreated group
What are the types of controls?
- geographic (patients with same disorder seen at another hospital without new intervention)
- historical (patients with same disorder seen in past before use of new intervention)
- randomised (allocations to group determined by chance)
What is the benefits of randomised controls?
Proper randomisation helps ensure groups receiving treatment A is similar to B
- avoids selection bias
- happens after groups determined eligible
What are the 2 types of blinding?
- single blind when patients don’t know what treatment they are on
- double blind when observers and patients do not know
What is a parallel group controlled trial?
When effect of a treatment is not reversible
- randomise individuals into 2 groups (treatment A and B), then record outcome for each
What is a crossover controlled trial?
When effect of treatment is reversible
- randomise individuals into 2 groups (treatment A and B), record outcome of each, put on other treatment then record outcome
What are the advantages of crossover trials?
- each patient has own control
- small sample size to get same observation numbers
- better for subjective measurements
What are the disadvantages of crossover trials?
- more time consuming
- carry over effects of one treatment (wash out period prevents this)
What are cluster randomised trials?
randomise pre-existing groups to 1 of 2 treatments
- avoids contamination
- enhances compliance
What are the phases of developing and evaluating a new drug?
Preclinical
Phase 0-4
What is the preclinical phase?
Non human study
- in vitro and in vivo animal experiments
- preliminary efficacy/toxicity/pharmacokinetic info
What is Phase 0?
First in-human trials
- small no. of subjects given sub therapeutic dose of drug to determine pharmacodynamics and kinetics
What is Phase 1?
Screening for safety
- test on healthy volunteers for dose ranging
- determine if safe to check for efficacy
What is Phase 2?
Assess efficacy and safety
- can it have a therapeutic effect
- design as case series or randomised controlled trial
What is Phase 3?
Assess efficacy and safety
- randomised controlled trial on large no. of patients determining therapeutic effect
What is Phase 4?
Post marketing surveillance
- safety surveillance
What is the advantage of registering clinical trials prospectively?
- assist planning new trials
- avoid unnecessary research duplications
- encourage research group collaboration
What is the relative risk?
risk of death in treatment group/risk of death in control group
risk of death in treatment group = no. of deaths in treatment group/no. of patients in treatment group
- same for risk of death in control group
- if no effect of treatment RR=1
- if RR is 0.86, treatment makes it 14% less likely to have a certain disease
What is intention to treat?
Comparison of all subjects based on treatment group assigned, regardless of compliance
- poor compliance = reduced ability to detect treatment difference
What is on treatment?
Comparison of subjects who took treatment
How is compliance maximised?
- selection of patients (not too ill)
- double blind design
- run in period where all get treatment
What is number needed to treat?
- average number of patients who need to treated to prevent additional bad outcome
What is absolute difference in risk?
= risk of death (control) - risk of death (treatment)
Why is it bad if the sample size is too small?
Study does not have enough power
What is meta-analysis?
Purpose to bring together all evidence to more powerfully estimate effect size
What are the issues in meta-analyses?
- heterogeneity (variation in study results)
- publication bias (studies with more favourable results more likely to be published)
