Lecture 2 - Randomised trials and real-world evidence

Question 1

Randomised Controlled Trials

Answer 1

Subjects randomised

Removes selection bias

Question 2

Selection bias

Answer 2

Association between exposure and health outcome is different for those who complete a study compared with those who are in the target population

Question 3

Matching

Answer 3

Construct intervention and comparison groups that are as similar to each other as possible

Question 4

Weighting

Answer 4

Assign intervention and control observations weights that balance both arms
Allows datasets to look the same

Question 5

Mean difference

Answer 5

Mental wellbeing, scale scores

Outcome variable should be a scale e.g. quality of life

Question 6

Risk Ratio

Answer 6

Risk of an event in one group divided by risk of event in another group

Question 7

Rate Ratio

Answer 7

Rate of an event in one group divided by rate of event in another group
Requires estimation of time at risk

Question 8

Person-years

Answer 8

total number of years subjects last in study

Question 9

Hazard Ratio

Answer 9

Time-to-event outcome
Instantaneous incidence rate in a group for those who have not accrued the outcome divided by the instantaneous incidence rate in another group for those who have accrued

Question 10

When is hazard ratio used

Answer 10

If event can only happen once

When the incidence of an outcome occurring in the control group is stable over time