Design and modeling of a computer-based medical decision analysis system for the MEDAS project.

摘要

Medical decision making in treatment and testing is known as one of the most critical tasks for physicians in medical care. Currently, physicians usually make clinical decisions based on their intuitive judgments. While this intuitive approach is suitable for those simple and fairly straightforward clinical situations, it is difficult and insufficient to handle clinical situations that are more ambiguous or complex, or situations that involve dangerous or costly treatment or tests. Therefore, a medical decision analysis system applying modern computer technology and formal medical decision analysis methodology to help physicians quickly and effectively solve this kind of problem is undoubtedly useful.;This thesis focuses on the design and modeling of a computer-based medical decision analysis system (MEDATT) for the MEDAS project. The main purpose of the system is to develop an easy-to-use, clear, and effective medical decision support tool to help practicing physicians in making important yet difficult decisions in treatment and testing. The fundamental methodology used in this system is based on Weinstein and Fineberg's method and the concept of the value of clinical information.;Three models in treatment decision analysis have been discussed--the automated consultation model, the user self-consultation model, and the training model. Each model provides a different function for the system. The automated consultation model serves as an expert system to automatically execute treatment decision analysis (risk, cost, life expectancy, patient morbidity, etc.) for the physician. The user self-consultation model performs treatment decision analysis for physicians based on their beliefs or assessments. The training model helps physicians acquire basic treatment decision analysis concepts and skills applied in the system.