Randomised controlled trials Robert West University College London

  • Slides: 5
Download presentation
Randomised controlled trials Robert West University College London

Randomised controlled trials Robert West University College London

Objectives You should be able to: 1. Describe what is an RCT 2. Explain

Objectives You should be able to: 1. Describe what is an RCT 2. Explain when is it appropriate to do an RCT 3. Design a simple RCT

What is an RCT? 1. 2. 3. 4. An experiment Involving a sample of

What is an RCT? 1. 2. 3. 4. An experiment Involving a sample of the population Randomly allocated to 2 or more groups Each group being delivered a different intervention (usually intervention versus control) 5. With one or more outcome measures obtained 6. To assess the effect of the intervention on the outcome(s) relative to each other

When is it appropriate to do an RCT? 1. When the research question can

When is it appropriate to do an RCT? 1. When the research question can usefully be answered by comparing only a small number of interventions 2. The interventions can be adequately constructed and implemented in a way that generalises beyond the study context 3. It is ethical and practicable to randomly allocate enough participants who are representative of the population of interest 4. There is unlikely to be substantial drop-out of participants after randomisation or during any follow-up period 5. Participants in all groups will be fully exposed to their respective intervention 6. Participants will not be exposed to the interventions applying to other groups 7. There is sufficient confidence that the intervention conditions will produce different results to merit the investment in time and resources

Starting to design an RCT 1. Research question 2. Study sample a. Inclusion and

Starting to design an RCT 1. Research question 2. Study sample a. Inclusion and exclusion criteria b. Recruitment and selection c. Retention 3. Intervention/control conditions 4. Outcome measure(s) a. Follow-up duration b. Choice of measure(s)