ABSTRACT
In medical research, a fundamental distinction can be made between observational and experimental studies. In experimental studies, investigators perform an action intended to interfere with the course of the disease or condition under study. The prototype is the randomized experiment, in which patients are randomly allocated to an experimental intervention and a comparator or a placebo. For example, patients with subclinical hypothyroidism (a condition in which thyroid function is slightly diminished) might be randomized between treatment with thyroid hormone substitute or an identical looking but inert placebo. With such an experiment, investigators could aim to answer the question of whether treatment improves cardiovascular risk associated with this condition. The patients would be treated for the purpose of the study, and arguably, many patients would not have been treated in the same way had they not been study participants (e.g., because their doctors do not think that their subclinical hypothyroidism importantly affected their health). The outcome of such an experimental study is a priori uncertain, and besides examples in which the treatment helped or made no discernible difference, there are examples of experimental treatments that turned out to worsen the disease course (Besselink et al. 2008).
