Evaluating Research

Question 1

Give an example of a drug which failed in market due to incorrect usage

Answer 1

Clomicalm - sales rocketed, vets handed out willy-nilly, not used properly - no results, sales plummeted

Question 2

What is the order of reliability of types of evidence?

Answer 2

Anecdotal (esp. drug reps will make this up!) 
Case reports 
Case series 
Cohort studies 
Controlled studies

Question 3

What is a case study?

Answer 3

Observing one animal over time

Question 4

What is a case series?

Answer 4

Multiple case studies

Question 5

What effects must be controlled for in an experimental study?

Answer 5

Individuals matched for age/breed/weight etc.
High numbers to account for individual variation
Order effects of the change implemented (randomise order)
Controls
Placebo effect
Confounding factors associated with the implemented change
Sampling bias of breed etc. - choose one breed first to show an effect, then generalise to whole population in subsequent studies

Question 6

What is reporting bias and publication bias?

Answer 6

Reporting bias - only positive findings reported (commercial, pre-registration)
Publishing bias - only positive results published (non-exciting ones not)

Question 7

How may the statistics of a study be affected?

Answer 7

Multiple things looked at, leading to adjusted P value being necessary
As opposed to hypothesis driven , where p value of <0.05 sufficient