Health Economics

  • Confounding occurs when an effect is attributed to an independent variable X when in fact it is due to an omitted (confounding) variable B, which is correlated with both X and the outcome (dependent variable) of interest. For example, higher socioeconomic status is strongly associated with both more frequent use of hormone replacement therapy and lower risk of coronary heart disease. The association between HRT and lower CHD has been erroneously interpreted as causal (Humphrey et al., 2002).