To use or not to use the odds ratio in epidemiologic analyses

摘要: This paper argues that the use of odds ratio parameter in epidemiology needs to be considered with a view specific study design and types exposure disease data at hand. Frequently, measure is being used instead risk or incidence-proportion cohort studies as an estimate for incidence-density case-referent studies. Therefore, analyses epidemiologic have produced biased estimates presentation results has been misleading. However, can relinquished effect these designs; and, application case-base sampling approach permits incidence difference measures estimated without any untenable assumptions. For Poisson regression, not interest; only rate are relevant. conditional logistic regression matched studies, remains useful, but when it interpreted ratio. Thus should, general, give way choice epidemiology.