Rewriting-Based Computer-Interpretable Clinical Practice Guidelines
Cyber Physical Systems
Clinical Practice Guidelines (CPGs) are systematically developed statements to assist healthcare professionals with decisions about specific clinical circumstances. Produced by domain experts, they they codify recommended treatments, and are often used as models for building Decision Support Systems (DSS) that assist healthcare professionals adhere to CPGs. CPGs are usually specified informally, or semi-formally (via flowcharts). This, however, may cause the behavior of DSSs to deviate from intended treatment in the corresponding CPGs, and introduce challenges in verifying correctness of the DSSs. In this paper, we propose a novel method of formally specifying CPGs by for- malizing them in K, a rewriting-based executable semantics framework . Our K-based guidelines are executable, allowing them to serve as correct-by-construction models in DSSs, and can be analyzed using tools in the K ecosystem. To support our claims, we present a DSS based on Advanced Cardiac Life Support (ACLS) CPG formalized in K, discuss challenges and present directions for future work.