Randomised Control Trials

Question 1

What are the types of intervention studies?

Answer 1

uncontrolled before and after study
controlled before and after study
RCT (individual vs cluster randomisation)

Question 2

In an uncontrolled before and after study, what other than the intervention might lead to change?

Answer 2

regression to the norm
seasonal change
policy change
practice change
measurement change

Question 3

Who might determine who is put into which group in a controlled before and after study?

Answer 3

participants (those selecting active treatment probably different to control group)

doctor (put more frail people into control)

trialist (might know who might do better in each arm)

Question 4

What are the 4 key features of an RCT?

Answer 4

randomisatoin
concealment of allocation
blinding
intention to treat analysis

Question 5

In a controlled before and after study, what other than the intervention might lead to change?

Answer 5

age
comorbidity
severity of illness
genetic difference

Question 6

Why must allocation be concealed in an RCT?

Answer 6

doctors may be able to guess the underlying sequence of allocation, and hence try to influence it (due to prognosis of patients or likelihood to adhere)

Patient Preference

Question 7

What are the 3 key stages to CASP framework for RCTs?

Answer 7

A - are the results of the trial valid?
B - what were the results?
C - will the results help locally?

Question 8

What 3 qualities must the outcome measure possess in an RCT?

Answer 8

reliable (ability to produce consistent, reproducible estimates of true effect)

valid (measures the construct that it claims to measure)

Responsiveness (can detect changes in the construct to be measured over time)

Question 9

What ethical issue is posed by randomisation?

Answer 9

individuals must not be disadvantaged by being randomised in intervention or control
we can only ethically randomise if we genuinely don’t know which is better (equipoise)

Question 10

What are the different types of control in an RCT?

Answer 10

usual care
no treatment (if there is no usual care)
placebo (helps blind people)
sham intervention (blinds patients)

Question 11

What is achieved by randomisation?

Answer 11

minimised the effect of confounding

Question 12

define confounder

Answer 12

a variable, other than the one studied, that can cause or prevent the outcome of interest

Question 13

What are the 3 qualities of a confounder in an RCT?

Answer 13

associated with outcome
not on causal pathway
distribution must differ between groups

Question 14

In an RCT what would the intervention ideally be?

Answer 14

all those allocated are offered it and adhere to it

none allocated to the control group get it, and they don’t seek alternative treatments to compensate

other than the intervention the groups are treated equally

Question 15

What 3 problems might arise with regards to the intervention in an RCT?

Answer 15

Non adherence (this is an important outcome!)

comparator group gets extra (performance bias)
they might get the treatment anyway (contamination)

intervention group gets extra (performance bias)

Question 16

What does blinding the participants achieve in an RCT?

Answer 16

avoids disappointment for ‘usual care’ group and means they don’t change their behaviour

Question 17

What does blinding the HCPs achieve in an RCT?

Answer 17

avoids any compensatory activities for the usual care group

avoids complementary treatments for intervention group

Question 18

What does blinding the researchers and statisticians achieve in an RCT?

Answer 18

avoids any bias in measuring or interpreting outcomes

Question 19

What are the different degrees of blinding in an RCT?

Answer 19

unblinded / open label
single blinded
double blinded
triple blinded

Question 20

What is single blinding?

Answer 20

participant is blinded

Question 21

What is double blinding?

Answer 21

participants and researchers / HCPs blinded

Question 22

What is triple blinding?

Answer 22

Participants, researchers / HCPs, and statistician blinded

Question 23

what is ITT?

Answer 23

intention to treat analysis

gives you the effect of assignment to the interventions at baseline, regardless of whether the interventions are received as intended

avoids allocation bias

Question 24

What is per-protocol analysis?

Answer 24

for an intervention where adherence is low, sometimes it is useful to estimate teh effect of adhering to the interventions as specific in the trial period

only participants who adhered are included

shouldn’t be in the main analysis and is usually exploratory

Question 25

How can you avoid detection bias?

Answer 25

assessing outcomes in the same way for both study arms

Question 26

What might study losses relate to?

Answer 26

outcome (sicker people in attrition bias)
allocation (intervention participants lost if they don’t like it or control participants lost if unblinded)
allocation and outcome (attrition bias)

Question 27

What questions might you ask to see if results can be applied locally?

Answer 27

would i be likely to see a similar effect if the intervention was delivered locally?

what was the population of the trial?

what is usual care like?

Question 28

When might you perform a cluster |RCT?

Answer 28

in an intervention with a high risk of contamination

when is requires a HCP to change their behaviour

is delivered in a situation where it wouldn’t be possible for some people to get one intervention and others something different