Confounding and Effect Modification in Research

An association that is not causal (Alternate explanations) is seen between outcome and exposure wheneverthe study has confounding factors, bias and random error/chance error. Bias in any study leads to systematicvariation of inferences or results or interpretations from the true picture. One of the most important biasencountered during research is confounding. Confounding is seen when a variable is observed to be associatedwith both the exposure and the outcome but is not a part of the causal pathway. Confounding factors, whenpresent, in any study, are the “Nuisance” and can be the cause in part or in full of the observed associationbetween the disease and exposure. Effect modification, on the other hand, is seen when various effects arebrought about among different subgroups by an exposure and this can be handled by doing stratification. It isthe outcome that is linked to the effect modification and not the exposure. This article will define and discussin detail confounding, and the concept of effect modification so that drawing conclusions of a study can bedone after considering these errors.