Quantitative - design, sampling and randomisation issues in randomised controlled trials Flashcards
What is the basic layout of a RCT?
study pps –> intervention group and control group –> outcome present in either group? (yes or no)
what is the design of a RCT?
- all trials have an intervention
- PPs randomised to either recieve the intervention or not (control group)
- sometimes there’s an attention arm (similar to intervention but without ‘active’ ingredient)
- outcomes of interest = measured
What are hypotheses?
- trials frquently start with them
- predictions about what they expect the intervention to achieve
What is a null hypothesis (Ho)?
- there will be no difference between the control and intervention arms
- assumed true at the start and has to be disproved
What are variables?
factors being investigated
What is a dependant variable?
-often the outcome of interest e.g. healing time
What is an independant variable?
-the intervention factor e.g. the dressing being used
How else can the IV and DV be described?
like cause (effect of dressing - IV) and effect (healing time - DV)
How is the population chosen in trial recruitment?
- target group researchers are interested in
- must be cleary defined (so know wwho results are applicable to)
- a sample will then be drawn (representative of the population)
What critera do trials have when recruiting?
inclusion and exclusion
What types of sampling are there?
probability and non-probability
What is probability sampling?
- gives an unbiased sample where everyone who meets the criteria has a chance of selection
- choose the sample of those entering the trial and is different to them being randomly allocated to groups once in the trial
What is non-probability sampling?
-non-random sampling and the chance of being selected cant be estimated
What are the four types of probability sampling?
simple random, stratified random, cluster and systematic random
What is simple random sampling?
- random selection of everyone in the population list
- rarely done because difficult to get population list
What is stratified random sampling?
-put in groups according to characteristics (lile gender) then randomly selected
What is cluster sampling?
-random selection of larger units (like hospitals) which PPs are then randomly selected from
What is systematic sampling?
-random selection at predetermined intervals
What are the factors affecting sample size?
- population factors - similarity of population to each other, expected rate of ‘event’, expected attrition (loss to study)
- design - number of variables, sampling
- measurement - sensitivity of measures
- practical factors - cost and convenience
What is a power calcuation and when is it performed?
- used to calculate the minimum effect sample size that is likely to be detected in a study using a given sample size.
- also used to calculate the minimum sample size required so that one can be reasonably likely to detect an effect of a given size
How are patients allocated to trial groups in RCTs?
-usually in a sequence over a period of time
How are patients allocated in to arms of the trial?
-methods of randomising them into the arms of the trial (intervention or control) are similar to the random selection of the sample
What are the three types of randomisation?
simple, block, stratified
What is simple randomisation?
-‘tossing a coin’ but usually done with random number tables
What is block randomisation?
-to keep numbers in each group close
What is stratified randomisation?
-to balance chosen characteristics across the arms of the trial
What does it mean when a trial is single blind?
- one person (usually PP) knows which arm of the trial thy are in (e.g. if its obvious what treatment they recieve)
- perosn assessing the outcome doesnt know
What does it mean when a trial is double blind?
neither PP or person assessing outcome knows the arm (e.g. placebo)