BMIR Research in Progress: Samson Tu & Mor Peleg “Decision Support for Multi Morbidities”

When:
May 18, 2017 @ 12:00 pm – 1:00 pm
2017-05-18T12:00:00-07:00
2017-05-18T13:00:00-07:00
Where:
MSOB, Conference Room X-275
1265 Welch Rd
Stanford, CA 94305
USA
Cost:
Free
Contact:
Marta Vitale-Soto

Samson Tu, MS,
Sr. Research Engineer,
Stanford Center for Biomedical Informatics Research,
Stanford University, &

Mor Peleg, PhD,
Visiting Associate Professor, Stanford University,
Dept. of Information Systems, University of Haifa

ABSTRACT:

Realizing the proven capabilities of computers in better processing of complex knowledge and data, medicine is heading toward computer-supported decision making. Following this vision, several research groups have developed languages that allow the representation of CPGs as computer-interpretable guidelines (CIGs). CIG engines match a patient’s data to the medical knowledge contained in the CIG in order to automatically deliver patient-specific recommendations at the point of care. However, existing CIG formalisms have not yet demonstrated in practice effective mechanisms for integrating CIGs to handle multi morbidities. The aim of this research is to develop and evaluate a new methodology for integrating the knowledge of CIGs for different chronic diseases to create non-conflicting management plans for patients with chronic comorbidities.

In the talk, I will present our methodology, which is a work in progress. Our methodology takes a goal-based approach that is modular and reusable. We envision a system of single-disease guidelines acting as agents whose invocation and conclusions are coordinated b a “Controller” agent that uses design patterns to recognize conflicts between different CPG goals and actions. The Controller then suggests ways in which these conflicts may be mitigated. It relies on CPG knowledge supplemented with general medical knowledge about the physiological effects of drugs and about drug hierarchies. Our approach is being implemented using existing commercial tools and standards. We use the PROforma CIG formalism and tools and the NDF-RT drug ontology for representing guideline and general medical knowledge, and the HL7 virtual medical record standard for structuring, storing, and retrieving patients’ data. The talk will focus on demonstrating how we have been developing our approach by applying information system analysis methods (Sequence Diagrams) to use-case patient scenarios of multi-morbidity patients.