The goal is a user-friendly software system with dual functions of intelligent tutor and expert clinical consultant in cancer pain management. Targeted users comprise medical, nursing, and pharmacy disciplines from student through practitioner levels. Features of the OPIAT approach include operability in PC environments; a visual, directly-controllable interface (windows, mouse, etc.); and run-time integration with CLIPS (a commercial data-driven, rule-based expert system shell). The domain of inquiry is subdivided into problems subdomains, viz., patient data input, pain diagnosis, management (intervention) decision, and simulation of probable outcomes to selected decision. The work plan includes: (1) intensive knowledge engineering (based on literature, consultants, and testing of evolutionary knowledge- base designs); (2) functional description of a comprehensive software package evolving from our experience with the Phase I prototype; (3) development (from the functional description) of a working OPIAT program, initially in two modes, separately, then combined into an integrated package: (4) extensive testing of the two modes by targeted users, with incorporation of feedback into a finalized package; and (5) software documentation (user's and program maintenance manuals.) The finalized package will be installed at the NIH Clinical Center to serve as a testbed for subsequent Phase III evaluation.
