Appraising trials Flashcards
1
Q
When looking at tables, what should you analyse? (6)
A
Title, headings, footer Multiple measures to use Design of experiment (placebo/PT) Raw data and statistical comparison Diff between groups at each stage Temporal relation for improvement
2
Q
Appraisal of a trial (7)
A
- Randomised
- Allocation sequence concealed
- Groups similar at baseline/start of trial
- Blinding (pts/therapists/assessors/study personnel)
- No missing data, complete follow-up
- ITT
- Role of chance (power calculation)
3
Q
Why conceal allocation?
A
Sequence could be manipulated
4
Q
Similar at baseline, what to consider? (2)
A
Influenced outcome/importance?
Researchers corrected for difference in analysis?
5
Q
If it is a subjective outcome measure, can a pt be blinded to group?
A
NO. They are “the assessor”.
6
Q
- Missing data, follow-up - what to look for? (2)
A
Why and how many left study.
> 15-20% loss to follow-up = caution
7
Q
What and why ITT? (1, 1)
A
Analyse as intended to be treated.
More realistic and preserves randomisation.
8
Q
Why 7. Power calculation?
A
Avoid Type II error
9
Q
What is a Type II error?
A
Failure to find and report a relationship when one actually exists)
