Randomised Controlled Trials

Question 1

Why use RCTs?

Answer 1

Most common type of study doctors use to inform treatment decisions

Question 2

What does RAMBOMAN stand for?

Answer 2

Recruitment
Allocation
Maintenance
BOM = Blind + Objective Measurement
ANalysis

Question 3

What questions should be considered for Recruitment?

Answer 3

Can you apply results based on these participants to your patients?
Source of EXTERNAL bias as usually volunteers + questions about motives

Question 4

What questions should be considered for Allocation?

Answer 4

Was a valid randomisation method used?
Were the exposure + control groups similar at baseline?
Randomisation may not work + must evaluate the method by which people were randomised

Question 5

What questions should be considered for Maintenance?

Answer 5

did participants remain in the study in the follow up time?
Did participants remain in their allocated groups?
More difficult in longer studies + when requires adherence

Question 6

What questions should be considered for Measurement?

Answer 6

were outcome measurements blind to the participants allocation status?
When outcomes subjective = concern that causes potential placebo effect (why need blinding)
Need to have both groups thinking they are getting the same

Question 7

What questions should be considered for analysis?

Answer 7

Were ITT analyses done?
ITT = takes account of everyone from the start of the study, and if people moved b/w groups not taken account of
OT = only measures those who remain in their allocated groups and potential for error from disruption of the randomisation process (confounding)

Question 8

What causes random error?

Answer 8

SAMPLE SIZE -
RCTs usually always too small due to high cost + difficulty with recruitment
usually use meta-analysis for reviewing the effectiveness of treatments

Question 9

What processes do you need for successful randomisation?

Answer 9

1. generating random number sequences to allocate

2. Concealing those involved from knowing their allocation

Question 10

What is quasi-randomisation?

Answer 10

any method that allows foreknowledge of the treatment that would be assigned e.g. allocating by age, initials, surnames

Question 11

What is stratified randomisation?

Answer 11

dividing participants into key groups first e.g. sex, and then randomising
Ensures balance between groups for “key” variables

Question 12

What is block randomisation?

Answer 12

Ensures numbers of participants are kept even between the groups
create blocks of sequences
2 study groups A + B and create blocks of 4 participants
possible sequences = ABBA, BAAB, ABAB, BABA, AABB, BBAA
then randomly choose sequences which ensures with each block you get 2 allocated to each group