Lecture #22 - Confounding Pt2

Question 1

1. What are the three ways you can control for confounding in study analysis?
2. Each can be used to control for what and evaluate what?
3. Must have measured what?

Answer 1

1. Stratification, multivariate analysis, and standardisation
2. Each can be used to control for confounding and evaluate whether it has occurred
3. Must have measured potential confounder (so need to plan how to measure them in study analysis phase)

Question 2

Stratification

1. What is it?
2. What is the basic recipe for this (4 steps)
3. What are the three pros of stratification?
4. What are the two downsides?

Answer 2

-

Question 3

Multivariate analysis

1. Statistical method for estimating measure of ass whilst….
2. Can work in situation when what won’t work?
3. What term are the variety of techniques recognisable by?
4. Good for when? (Kinda answering number 2)

Answer 3

-

Question 4

Standardisation

1. Two times when you do this - what are they?
2. Other than age, what else could you use this for?
3. What are the two “But” points?

Answer 4

-

Question 5

Miscellaneous points:

1. What can you say about residual confounding?
2. Can’t control what you haven’t measured - explain.
3. What is the importance of measuring potential confounding?

Answer 5

-

Question 6

Matched analysis:

1. What do you have to do with matched case control studies?
2. What will happen to the estimate of measure of association if you don’t do this analysis for matched case control studies?
3. What is this not a form of?

Answer 6

-

Question 7

Assessing confounding:

1. So need to be controlling for potential confounders in analysis. What can we evaluate the magnitude of confounders through?
2. How much change is indicative of confounding - what is the gnawed all guideline (ie how much can the crude and the adjusted differ?)

Answer 7

-

Question 8

Effect modification:

1. “The association between exposure and outcome differs across _____ of the effect modifier”
2. Whereabouts on the flowchart will you get to find out whether it’s confounding or effect modification at play?
3. How should you report finding if you find its an effect modifier?
4. The next lecture - what’s a simple explanation of effect modifier?

Answer 8

-