Details

Autor(en) / Beteiligte

• Yang, Junyi
• Xiao, Liang
• Li, Kangning

Titel

Modelling clinical experience data as an evidence for patient-oriented decision support

Ist Teil von

• BMC medical informatics and decision making, 2020-07, Vol.20 (Suppl 3), p.138-138, Article 138

Ort / Verlag

England: BioMed Central Ltd

Erscheinungsjahr

2020

Link zum Volltext

Quelle

MEDLINE

Beschreibungen/Notizen

• Evidence-based Clinical Decision Support Systems (CDSSs) usually obtain clinical evidences from randomized controlled trials based on coarse-grained groups. Individuals who are beyond the scope of the original trials cannot be accurately and objectively supported. Also, patients' opinions and preferences towards the health care delivered to them have rarely been considered. In this regards, we propose to use clinical experience data as an evidence to support patient-oriented decision-making. The experience data of similar patients from social networks as subjective evidence and the argumentation rules derived from clinical guidelines as objective evidence are combined to support decision making together. They are integrated into a comprehensive decision support architecture. The patient reviews are crawled from social networks and sentimentally analyzed to become structured data which are mapped to the Clinical Sentiment Ontology (CSO). This is used to build a Patient Experience Knowledge Base (PEKB) that can complement the original clinical guidelines. An Experience Inference Engine (EIE) is developed to match similar experience cases from both patient preference features and patient conditions and ultimately, comprehensive clinical recommendations are generated. A prototype system is designed and implemented to show the feasibility of the decision support architecture. The system allows patients and domain experts to easily explore various choices and trade-offs via modifying attribute values to select the most appropriate decisions. The integrated decision support architecture built is generic to solving a wide range of clinical problems. This will lead to better-informed clinical decisions and ultimately improved patient care.