what are the advantages of randomised control trials?
- intervention and control groups will be similar in all respects (except the intervention), minimising selection bias and confounding.
- if the participants are blind to the treatment allocation, reporting bias can be minimised
- if the investigators are blind to the allocation, observer bias can be minimised
- multiple outcomes can be examined
- the incidence rates of the outcome can be measured
disadvantages of randomised control trials?
- expensive to conduct
- intervention studies are impossible to conduct for ethical/logistical reasons in some situations
- time-consuming
what are the main components of a randomised control trial?
population
intervention
comparator
outcome
what is restricted randomisation?
refers to any procedure used with random assignment to achieve balance between study groups in size or baseline characteristics
what is block randomisation?
computer generates complete blocks
ensures that the number of participants allocated to each group is equal after every block of x patients has entered the trial
what is stratified randomisation?
useful if other risk factors have a strong influence on the outcome
what is allocation concealment?
refers to making sure that the method of allocating a participant to a group can’t be revealed
(e.g. intervention, placebo, intervention, placebo…)
what are the advantages of parallel group trials?
- easily applicable and understood
- multiple groups can be studied
- easier to multi-centre studies
what are the disadvantages of parallel group trials?
- larger sample size
- more expensive
- with multiple groups may be difficult to analyse
what are the advantages of crossover trials?
- fewer patients needed
- eliminates between patient variability
- test is within-patient, so balanced
- can reduce ethical issues with placebo
what are the disadvantages of crossover trials?
- carry-over effects possible
- cant be used in curable treatments or for long-term treatment
- data wasted when patients drop out in first period
- duration of trial longer
what are the advantages of factorial trials?
can assess combined effects in one trial
what are the disadvantages of factorial trials?
- interaction effects
- more difficult to meet inclusion criteria
- complex
what is cluster randomisation?
when people are organised in natural groups (i.e. clusters)
clusters are randomised, not individuals
what are the disadvantages of cluster randomisation?
- need to randomise many clusters
- need increased sample size
- subjects within a cluster may be different to subjects in other clusters
- allocation concealment is more difficult
- analysis is more complicated
what is performance bias?
systematic differences in the care provided to the participants in the comparison groups other than the intervention under investigation.
often intervention receive more attention that the controls, particularly if not blinded.
