Date of Award

Winter 2016

Document type

Thesis

Degree Name

PhD (Doctor of Philosophy)

First Supervisor

Professor Tom Fahey

Second Supervisor

Dr Damon Berry

Funder/Sponsor

TRANSFoRm is partially funded by the European Commission – DG INFSO (FP7 247787). This work was partially funded by the Health Research Board (HRB) of Ireland through the HRB Centre for Primary Care Research under Grant HRC/2007/1

Keywords

Clinical Prediction Rules, Clinical Decision Support, Electronic Health Records, Knowledge Representation, Ontologies, Health Informatics

Abstract

Diagnostic error is a threat to patient safety in the context of primary care. Clinical prediction rules (CPRs) are a form of structured evidence based guideline that aim to assist clinical reasoning through the application of empirically quantified evidence to evaluate patient cases. Their acceptance in clinical practice has been hindered by literature-based dissemination and doubts regarding their wider applicability. The use of CPRs as part of electronic decision support tools has also lacked acceptance for many reasons: poor integration with electronic health records and clinician workflow, generalised guidelines lacking patient-specific recommendations at point-of-care, static rule based evidence that lacks transparency and use of proprietary technical standards hindering interoperability.

The ‘learning health system’ (LHS) describes a distributed technology based infrastructure to generate computable clinical evidence and efficiently disseminate it into clinical practice. This research describes an LHS based on computable CPRs for diagnostic decision support that makes use of aggregated sources of primary care electronic health record data to derive and disseminate computable CPRs.

Based on a literature review of clinical and technical best practice regarding use of CPRs, a theoretical model for CPRs supporting two critical aspects for a successful LHS is proposed: the model representation and translation of clinical evidence into effective practice, and the generation of curated clinical evidence that can dynamically populate those models thus closing the learning health system loop. A functional implementation of the theoretical model demonstrates an infrastructure that is model-driven, service oriented, constructed using open standards, and supports a learning evidence base derived from electronic sources of patient data.

A number of challenges exist for the LHS community to consider including medico-legal responsibility for generated diagnostic evidence, developing trust in the LHS, constraints imposed by clinical terminologies on evidence generation, and quality and bias of underlying EHR data for evidence generation.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 4.0 License.

File Size

11.2 MB

Comments

A thesis submitted for the degree of Doctor of Philosophy from the Royal College of Surgeons in Ireland in 2016.

Share

COinS