Misclassification of a confounder leads to a loss of ability to control confounding,
although control may still be useful provided that misclassification of the confounder is non-deferential. In general, control of confounding requires careful use of a prior knowledge, as well as inference from the observed data. Control in study design Confounding can be controlled in the study design, or in the analysis, or both. Control at the design stage involves three mains methods : 1. Randomization The example for randomization methods is random allocation that to exposure categories,but this is rarely an option in epidemiology which generally involves observational studies,but the method is debatable whether randomized studies are part of epidemiologi. 2. Restriction Restricting the study to narrow ranges (white males aged 34-54) has conceptual and computational advantages. But restriction can also decrease the potential study subjects such as younger and older people because they will not be observable. 3. Matching study on potentional confounders A third method of control involves matching study subjects on potential confounders. This will prevent age-sex-race confounding in a cohort study, but it seldom dont because it may be very expensive. Matching may actually reduce precision in a case control study if it is done on a factor which is associated with exposure, but is not a risk factor for the disease of interest. However, matching on a strong risk factor will usually increase the precision of effect estimate. Control in the analysis The analysis ideally should control simultaneously for all confounding factors. It is usually not possible to control simultaneously for more than 2 or 3 confounders in a stratified analysis. Control of confounding in the analysis involves stratifying the data according to the levels of confounders and calculating and effect estimate which summarizes the association across strata of confounders.