Advertisement

Methods to Control for Confounding

  • Bryan Kestenbaum
Chapter

Learning Objectives

  1. 1.
    Methods used to control for confounding include:
    1. a.

      Restriction

       
    2. b.

      Stratification

       
    3. c.

      Matching

       
    4. d.

      Regression

       
    5. e.

      Randomization

       
     
  2. 2.

    Restriction can be a powerful method to address a limited number of confounders.

     
  3. 3.

    Restriction by the indication for a drug can be used to address confounding-by-indication in observational studies of medication use.

     
  4. 4.

    Stratification involves dividing the study population into strata, and then weighing and combining the stratum-specific results.

     
  5. 5.

    Restriction, stratification, and matching may not be appropriate for dealing with multiple confounding variables.

     
  6. 6.

    Randomization balances both measured and unmeasured characteristics.

     
  7. 7.

    A factor is likely to be confounding a given association if adjustment for that factor substantially changes the strength of the association.

     

Keywords

Relative Risk Ghrelin Level Potential Confounding Variable Plasma Ghrelin Level Summary Relative Risk 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Copyright information

© Springer Science+Business Media, LLC 2009

Authors and Affiliations

  1. 1.University of WashingtonSeattleUSA

Personalised recommendations