Korean J Epidemiol.
1999 Dec;21(2):248-253.
A Review Study on Confounding Effect: Case-control Study
- Affiliations
-
- 1Biostatistics unit, Samsung Biomedical Research Institute.
- 2Department of preventive medicine, School of Medicine, Ajou University.
Abstract
-
Confounding is the distortion of a disease/exposure association brought about by other factors which are not considered in the study design or the data analysis. These factors are called confounding factors. We should be cautious in data analysis of observational study of association of disease/exposure, since confounding often occurred in observational study. This study examines confounding effect according to data pattern (the ratio of controls to cases, the ratio of exposures to non-exposures for each level of confounding factor), criteria for treating a variable as a confounding variable, and some notes for the analysis in case-control study.