randomised control trials Flashcards
What is the classification of the randomised control trial
Interventonal prospective analytic epidemilolgical study
trial and control arms are assigned
What is randomization
Random assignment of groups with equal probability of being in each
Goal: elimentant confounding increase confidence that association is due to exposure
Large enough sample -> confounding equal in eaching group, doesnt matter is we know about it or not as passively balances
What is the difference between randomization and random selection
Random selection = picking a random sample from population
randomization = random division between control and trial arms
What factors are important to maintain randomization
Concellment of allocation - dont when control/trial will be selected, prevents bias from reserchers of participants
methods to deal with loss to followup:
-Intent to treat - analyse as randomized, even if dropped out/stoped, to protect randomization
-pre-protocal - analyse as treated, more appropriate for efficeny trial rather than assocaition
Large numbers
What factors are importnat to reduce and lead to bias in an RCT
Blinding - Protection from bias, use placebo for control etc, blinding from all parties ideal
Non-adhearance - people who change their regimant throughout the trial
Loss to followup - impact results
What are two factors to mean that a RCT would be a good study design
Clinical equipoise = genuine uncertainty about the result
Exposure needs to be modifiable = something you can change
What are the strengths of an RTC
Best method to evaluate intervention
Strong for eveluating association
calcuate incidence
What are the weaknesses of an RTC
Slow and expensive
highly selective - can modify the genralisation
clinical equipose - genuine uncertancy about results
modifiable exposure required
