07 - 08 Randomized Controlled Trails (RCT)

Question 1

What is a ideal study design?

Answer 1

Outcome can only be explained with a causal effect and it’s random

Question 2

What is important to consider when designing a clinical study?

Answer 2

1. protocol and ethics
2. randomisation of intervention and control group
3. blinding
4. analysis by unbiased researchers
5. replication
6. benefits and risks

control: placebo/best available
representative patients
use relevant endpoint
superiority

Question 3

Define randomised controlled trial

Answer 3

• epidemiological experiment
• randomly chosen subjects from population
• study and control group
• rigorous comparison

can be preventive (e.g. vaccine) or therapeutic (e.g. new drug)

Question 4

What is a control group for?

Answer 4

allow to distinguish between outcomes from intervention or natural course or bias

Question 5

Explain structural, operational and observational equality and how it is archived.

Answer 5

• structural equality: compatibility between groups -> randomisation
• operational equality: avoidance of different behaviour because of expectation -> blinding
• observational equality: avoidance of different behaviour because of expectation -> blinding

Question 6

Name possible randomisations

Answer 6

• unrestricted (like throwing a dice)
• restricted:
  
  • blocked: avoid different groupsize -> random order of ‚blocks‘
  • stratified: avoid imbalances -> make (weight/age) groups before randomising those

Question 7

Name the possible types of blinding

Answer 7

• open (not blinded)
• single blind: physician knows if patient is under treatment/placebo
• double blind: no knows who is under treatment/placebo
• triple blind: like double blind + analysts don‘t know who is what group

Question 8

When is RCT justifiable?

Answer 8

• treatment is modifiable and can be controlled
• expected benefit is greater than harm
• uncertainty principle: physician doesn‘t know which treatment is better -> substantially unclear
• informed consent
• voluntary participation

Question 9

Name members of an ethical review board

Answer 9

• Physician
• Pharmacists
• Judges
• Ethicians
• Theologians
• Statisticians
• lay persons

Question 10

Explain Intention to Treat (ITT) and per protocol analysis

Answer 10

ITT:
- include all study participants (also drop outs and non-complained ones)
- maintains structural equality
- effect will be underestimated
preferable

PP analysis:
- use data according to treatment
- loss of structural equality (randomisation) possible

non-complainant -> swapped group or got completely different treatment

Question 11

Name the three major study endpoints

Answer 11

example: sun exposure as cause for skin melanoma
- intermediary outcome
sun burn
- proxy/surrogate parameter
referral to dermatologist because of suspected melanoma
- clinically relevant endpoint
melanoma diagnosis

Question 12

(Dis-)advantages of RCTs

Answer 12

Advantages:
- randomisation leads to structural equality
- strictest method to test hypotheses in epidemiology
- no chance to influence the exposure status

Disadvantages:
- complex and costly
- ethical questions
- relationship between patient and physician affected?
- if participants don‘t represent population of interest findings can‘t be generalised

Question 13

Placebo and Nocebo effect

Answer 13

Placebo -> intervention effect with placebo
Nocebo -> adverse intervention effect with placebo

Question 14

Special RCT designs

Answer 14

Sequential study design:
- interim analysis -> can/should study be continued?

Factorial study design:
- combination of several treatments

Cross-over study:
- one individual gets one after the other treatment