Comparing with 'Care as Usual' can be problematic

Randomized trials aim to estimate causal associations. To do that, the interventions (or conditions) to be compared need to be defined precisely. Hernan and Robins call this "consistency" of the counterfactual outcomes. The comparison of some (new) treatment to 'Usual Care' can be problematic if 'Usual Care' is ill-defined, as is often the case. The US PCORI institute has formulated requirements for justification and description of Usual Care. This blog post text addresses this issue and the inspirations that various fields of science took from Isaac Newton.