Stochastic Dominance and Medical Decision Making
|Speaker:||Moshe Leshno, Tel Aviv University|
|Date:||Wednesday 13 February 2013|
|Location:||Matrix Lecture Theatre, Building One|
Stochastic Dominance (SD) criteria are decision making tools which allow us to choose among various strategies with only partial information on the decision makers’ preferences. The notion of Stochastic Dominance has been extensively employed and developed in the area of economics, finance, agriculture, statistics, and marketing since the late 1960s.
Clinical guidelines which are based on the best available evidence are recommendations on the appropriate treatment and care of people with specific diseases and conditions. Guidelines help healthcare professionals in their work.
SD may tell us which of two medical treatments is preferred in the absence of full information on the patients’ preferences. Thus clinical guidelines may be evaluated by SD paradigm. The limitation of SD is that while “most” decision makers may prefer one uncertain prospect over another, SD rules, will not reveal this preference due to some extreme utility functions in the case of even a very small violation of these rules. Such strict rules relate to “all” utility functions in a given class including extreme ones which presumably rarely represents decision makers’ preference.
In this presentation we present a short review of the concept of SD and almost stochastic dominance (ASD) rules which formally reveal a preference for “most” decision makers, but not for “all” of them. We demonstrate how the SD criteria may be employed in medical decision making, using the case of small abdominal aortic aneurysms and screening for prostate cancer as an illustration.