Stochastic dominance and medical decision making

Area of Science:

• Decision Sciences
• Medical Economics
• Statistics

Background:

• Stochastic Dominance (SD) criteria are established decision-making tools used since the 1960s across various fields like economics, finance, and operations research.
• These criteria facilitate choices among strategies with incomplete information on decision-maker preferences, particularly useful for uncertain outcomes in areas like medical treatments.

Purpose of the Study:

• To review the Stochastic Dominance (SD) paradigm.
• To demonstrate the application of SD criteria in medical decision-making.
• To illustrate the use of SD with the specific case of small abdominal aortic aneurysms.

Main Methods:

• Review of the Stochastic Dominance (SD) paradigm.
• Application of second-degree stochastic dominance (assuming risk aversion) to categorize medical treatments into efficient and inefficient sets.
• Utilization of Prospect Stochastic Dominance (PSD) for divisions corresponding to S-shaped utility functions.

Main Results:

• Second-degree stochastic dominance can identify an efficient set of treatments that decision-makers should prefer, assuming risk aversion.
• Prospect Stochastic Dominance (PSD) allows for similar categorization applicable to a broader range of S-shaped utility functions.
• The study illustrates how SD can aid in selecting optimal medical strategies for conditions like small abdominal aortic aneurysms.

Conclusions:

• Stochastic Dominance (SD) criteria offer a valuable framework for medical decision-making under uncertainty.
• SD methods, including second-degree SD and PSD, can effectively guide the selection of superior treatment options.
• The application to abdominal aortic aneurysms highlights the practical utility of SD in clinical contexts.