38. RANDOMISED CONTROL TRIALS Flashcards
- What happens in a Randomised Control Trial?
- participants are assigned to a study group
THE PROCEDURES ARE CONTROLLED:
- they make sure that all the participants in the study are
treated the same
- except for the intervention (independent variable)
THE PARTICIPANTS ARE ASSIGNED TO SPECIFIC TREATMENTS AT RANDOM:
- this means that they have an equal probability of being
assigned to any group
- What are the groups that are being studies in RCTs called?
- they are called Study Arms
- they can also be called Treatment Conditions
- Looking at this example, what can we say about the Crude Odds Ratio?
- it is less than 1
- this means that the exposure has a protective effect
- Looking at this example, what can we say about the Adjusted Odds Ratio?
- it includes 1 in its value
- this means that there is no association
- What is the effect of Randomisation?
- it ensures that all the exposures in the sample as
evenly distributed across the groups - there is a control group for the known and the
unknown confounders
- What is the advantage of Randomisation?
- it gives the researchers a sense of confidence
- that the differences in outcome between the treatment
and the control - are a result of the treatment alone
- the test is fair
RANDOMISATION:
- will equalise the groups on all the other variables
- only the independent variable will have an effect
- Define: Allocation Concealment.
- the researcher who has generates the Random
assignment of exposures to participants - REMAINS BLIND
- to what kind of condition the participants will enter
- What happens when the Allocation is not concealed?
- the research staff are prone to assign patients who
performed better to the intervention - rather than assigning the control
THIS CAN BIAS THE TREATMENT EFFECT:
- up to 20 to 30 %
- What are the two main aspects of Randomised Controlled Trials?
- RANDOMISATION
- this reduces the risk of Confounding - ALLOCATION CONCEALMENT
- this has to do with Assigning
- this reduces the risk of Selection Bias
- What is Blinding?
- this prevents the Researchers from knowing certain
information about a participant - this prevents the information from affecting how the
researcher collects the data
- What is the best way to minimise Information Bias?
- Double Blinding
BOTH THE PARTICIPANTS AND THE RESEARCHER:
- are kept blind to what exposure the patient has
randomly been assigned
- What is an example of a negative effect of lack of Blinding?
- A PARTICIPANT
- may believe that a new drug is harmful
- they will self-report in an exaggerated and biased way - A RESEARCHER
- may believe that a drug is extremely beneficial
- they can over state the benefits of the drug
IN BOTH OF THESE CASES:
- the results differ from reality
- What is the set up of the simplest Randomised Controlled Trial?
IT HAS ONE INTERVENTION GROUP:
- this is the group receiving the treatment of interest
IT HAS ONE CONTROL GROUP:
- they are receiving the placebo
- or they are receiving another treatment
- Define: Placebo.
- this is a substance
- it is identical to the intervention substance
- it lacks the active ingredient
- this means it is chemically inert
- it has no physical effect on the participant
NB:
- this is what leads to the Placebo Effect
- What is the Placebo effect?
- it is a beneficial and psychological effect
- What causes the Placebo effect?
- THE PARTICIPANT’S BENEFICIAL RESPONSE TO THE
INVESTIGATION
- the patient’s response to a Therapeutic trial - THE PATIENT’S BENEFICIAL RESPONSE TO
OBSERVATION
- and their response to assessment - THE PATIENT’S BENEFICIAL RESPONSE TO THE
DOCTOR-PATIENT INTERACTION
- What do Placebo-Controlled Trials allow for?
- they allow for us to study the effectiveness of the
intervention treatment - it looks beyond all placebo effects
- What is the advantage of Double-Blinding?
- it reduces the influence of the expectations that the
participants have - it reduces the influence of the expectations that the
research staff have
THIS EXPECTATION IS ABOUT:
- which treatment will have a better effect on the
outcome
- What is the most feasible kind of Blinding for Drug Trials?
- Double-Blinding
- the effect of the medication is being compared to a
Placebo
- How can we alter the Placebo to reduce Chance?
- we can mask the Placebo
- so that it looks identical to the drug being tested
EXAMPLE:
- make it look the same
- make it the same size
- make it the same colour
- make it taste the same
- make it have some similar side effects
- In more complex RCTs, we can have several intervention groups and no placebo groups.
What is always necessary in the Analysis Stage of a RCT?
- one intervention group needs to act as the Control
Group
- What happens when we compare a new treatment to an old (conventional) treatment?
- we use the old treatment as the Control group
- What group is the Control when we are dealing with different dosages of a single treatment?
- the control group is the one with the lowest dose
- What is Partial Blinding ?
- this is the type of blinding we use when Double
Blinding is not possible
THE PARTICIPANT:
- is aware of what treatment they are receiving
THE RESEARCHER:
- is blind to the information about which treatment was
assigned to which participant
EVEN IF THE TREATMENT ASSIGNMENT IS KNOWN BY THE RESEARCHER THAT DELIVERS IT TO THE PARTICIPANT:
- the staff who assess the outcome should be kept blind
to this information
- What are studies in which we cannot make use of Double Blinding?
- studies that look at Behavioural Treatment
- What are Unblinded Trials?
- these are trials where neither the participant or the
researcher are blinded
IN THIS CASE:
- the participants and the researchers have to have the
same expectations
- about the control condition and the treatment
- Define: the State of Equipoise.
- it is the ethical justification of conducting a trial
- it is used when we do not know which intervention
condition will work best
- What does Randomisation prevent?
- it prevents Biased Assessment of the outcomes after
Randomisation - it prevents Measurement Bias
- it prevents Ascertainment Bias
- What does Blinding reduce?
- it reduces non-compliance
- it reduces contamination
- it reduces self-report biases
- it reduces investigator bias based on their opinions
- In which situations is Blinding not possible?
- in Open Trials
IN THIS CASE:
- a hard outcome must be chosen
- this is an exact numerical outcome
- the treatments must be standardised as much as
possible
- When is Blinding hard to maintain?
- when a drug shows specific side effects
- What is the difference between Allocation Concealment and Blinding?
ALLOCATION CONCEALMENT:
- takes place before the participants are allocated their
exposure
BLINDING:
- takes place after the participants have been allocated a
specific exposure
- it can take place during the study
- it can take place when the outcome results are being
collected
