The Precise Construction of Patient Protocols: Modeling, Simulation and Analysis of Computer Interpretable Guidelines

Abstract

Standardizing the care of patients with complex problems in hospital settings is a difficult challenge for physicians, nurses and other medical professionals. Diverse conditions further complicate patient management. While in acute care settings such as intensive care units, the inherent problems of stabilizing and improving vital patient parameters is further complicated by the division of responsibilities among different individuals and teams, in outpatient settings the management of chronic diseases introduces additional complications related to the long-term treatment of patients. The use of evidence-based guidelines for managing complex clinical problems has become the standard of practice. Computerized support for implementing such guidelines has tremendous potential; however, addressing this problem requires a carefully coordinated use of various techniques from the field of computer science, as guidelines developed by the medical community are not directly interpretable by computers. In this thesis, first, we present a survey of literature and a study on the open questions from the field of clinical decision support focusing on the use of model-based techniques for specifying and implementing evidence-based guidelines. Following the survey, we describe a model-based architecture for enabling the construction, management, verification and execution of such guidelines. The presented architecture is model-based in the sense that it relies upon the formal modeling of medical guidelines, including the specification of input parameters such as signs and symptoms, output parameters such as medical actions, and other guideline-related constraints such as rules, regulations and policies. The behavioral semantics of these models is provided by the application of custom-built formal behavioral templates defined with the help of Matlab Simulink/Stateflow and model composition. The benefits of our approach are illustrated with the modeling, execution and formal analysis of a clinically relevant example, a sepsis management guideline.