Exercise: Statins, pneumonia and confounding, 1

Question 1: What is meant by "adjustment for confounding"?

Suggested answer

Death or admission to an intensive care unit is the outcome variable. This may be related to several predictors, not only statins but also age, sex, blood pressure, cholesterol, condition of arteries, etc.

If two predictor variables are related to one another as well as to the outcome, their effects are said to be confounded, because the effects of each of them will include some of the effect of the other.

By "adjustment", we mean that we estimate the effects of one of these variables for subjects who have identical values of the other variable.

If the relationship between the the two predictors is strong, it becomes difficult to estimate the coefficients for both together and these estimates have large standard errors and wide confidence intervals. This is because a change in one coefficient can be compensted for by an opposite change in the other coefficient. A lot of possible pairs of values for the coefficients will predict the outcome with similar accuracy.

Back to Exercise: Statins, pneumonia and confounding.

To Applied Biostatistics index.

To Martin Bland's M.Sc. index.

To Martin Bland's home page.

This page maintained by Martin Bland.
Last updated: 12 December, 2006.

Back to top.