RCTs

Question 1

External validity define

Answer 1

Extent to which trial results can be generalised to population

Question 2

Describe how treatment preference would affect internal and external validity?

Answer 2

Impact on recruitment? Patients aware of treatments. Volunteer bias? This would threaten external validity if only volunteering because they are getting new drug.
Patients would be able to tell what drug they are getting if their symptoms differed from normal, then they would know they are in study but might stay just for sake of it. This would threaten internal validity.

Question 3

What is internal validity?

Answer 3

Extent to which differences in outcome can be ascribed to differences in treatment (rather than differences in group characteristics)

Question 4

How can results be affected by what the patient thinks?

Answer 4

Pt recieves preferred treatment:
- Potential therapeutic effect 
- Enhanced motivation, greater compliance
- Outcomes: ENHANCED REPORTING

Pt receives undesired treatment:
- Resentful demoralisation 
- Reduced accuracy on reported outcome 
- Drops out

Question 5

What are the advantages of patient preference trial?

Answer 5

• Good for mental, physical wellbeing
• Resentful demoralisation reduced
• Reduced drop outs
• Improvement in reported outcomes
• External validity is actually enhanced
• Randomised groups enhanced internal validity
• Comparison of randomised vs preference groups could fix it
• More life-like
  Methodological issues - more patients needed to do a stats trial where everyone gets what they want
• Would lead to death of statistical beauty that is RCT

Question 6

Problem with RCTs

Answer 6

Considered the gold standard. However, most RCTs “suffer” from selection bias, plus recruitment and retention difficulties

Question 7

External validity depends on characteristics of sample (participants) recruited – and whether they represent population they are meant to.

Will depend on recruitment, dictated by (5 points):

Answer 7

A) Characteristics of GP practices - different from others in population?
B) Patients selected - selection bias (systematic difference between participants and population they are meant to represent)
C) Volunteer bias – systematic bias between those who volunteer and those who don’t. If large number refuse to take part, external validity threatened
D) Recruitment - may be based on clinician choice
E) Patient treatment preference

Question 8

Why are RCTs considered gold standard? 3 points

Answer 8

1. Random allocation (minimising confounding)
2. Control group (allows comparison)
3. Double-blinding (where possible) to reduce assessor- and response-biases