CAT RCT

Question 1

Did the study answer a clearly focussed question?

Answer 1

Was it focused in terms of:
PICO

Question 2

How were participants randomised?
Was this efficient?

Answer 2

Aim:
* To ensure groups are similar across all patient characteristics
* to prevent over/under-esimtating effects
* to prevent selection bias
Methods of randomisation:
* Complete - use of random number generator
* +/- Stratification - with powerful confounders (age/sex) patients are stratified into different groups BEFORE randomisation so there are the same number of patients with/without the confounder in each arm of the trial.
* Minimisation: use of an algortith
* Replacement

Question 3

Were all patients who entered the study accounted for at its conclusion?

Answer 3

Consider:
* Losses to follow up - were they acknowledged?
* For ITT analysis the participants must be analysed in groups which they were randomised to (gives pragmatic estimate; better approximation of real clinical practice, more accurate picutre)
* The alternative fof ITT is per protocol analysis
* If study stopped early - why?

Question 4

Were participants in all groups followed up and data collected in same way?

Answer 4

• Follow up** intervals** - same timing?
• Follow up methods - in person/by phone: could result in measurement bias.
• Were **interventions similar **across study groups?
• Same attention from researchers and health workers
  *

Question 5

Did the study have enough participants to minimise the play of chance?

Answer 5

• Power calculation:
  1. Estimates how many participants are required to ensure the trial can identify a true null result: aka how many participants are needed to avoid a type 2 error (accepting the null wrongly)
  2. Important to assess whether statistical testing has taken place without adjustment of the p value (type 1 error: rejecting the null wrongly)