Medin: An expert system for processing medical insurance claims

摘要

The develoment of an expert system prototype is described. The system is intended to assist nurses who process medical insurance claims. The problem is essentially one of classification, where a set of codes and data on the claim form are used to arrive at a decision. We emphasize the lessons learned from knowledge acquisition and implementation and discuss the adequacy of rulebased systems for solving similar problems, including a discussion of how the prototype, although functional in itself, can be extended to become a fully viable system. In this article we detail the process of building the prototype. In particular, the knowledge elicitation stage—its duration, getting the expert to verbalize her information processing, and generalization from the set of examples available. We also highlight the benefits that insurance companies can derive from similar development efforts. We present the pros and cons of using shell-based packages for similar problems, including their convenient and user-friendly environment.