• Ontologies to Improve Chronic Disease Management Research and Quality Improvement studies – a Conceptual Framework [Presentation]

 Report from the EFMI PCI WG:

Workshop on using ontologies to define better chronic disease management research and quality improvement studies

EFMI Ontology Toolkit Survey 2012 – Round 1 – Online version 

EFMI Ontology Toolkit Survey 2012 – Round 2 – Online version 

Background – What is an Ontology?

Ontologies are important in informatics because they form a structured way that we can explore concepts and their relationships.

The commonly-accepted definition is:

An ontology is a specification of a conceptualization – Gruber (1992)



Developing a consensus about using ontologies to design research and quality improvement (QI) studies in primary and integrated care

Consensus development:

The consensus development process involved two rounds.  The first is an inclusive round – seeking to identify elements of a new framework and round two is intended to be selective – looking to achieve consensus about what is necessary.

Round 1:

Identify elements of a toolkit to support the use of ontologies in research or QI

   – This will be an inclusive list

Round 2:

To achieve a consensus, about which elements of the viagra overnite toolkit are essential

   –  Using the RAND Europe method


 1. Workshop synopsis from the proceedings of MIE2012

    Download workshop plan

de Lusignan S, Liaw S-T, Rahimi A, Poh N, Jones S.   Improving the Design of Research and Quality Improvement Using Routine Data in Chronic Disease: Ontology Driven Approach.  24th International Conference of the European Federation for Medical Informatics Quality of Life through Quality of Information – J. Mantas et al. (Eds.) MIE2012 / CD / Working Group Workshop

2. Presentation from the workshop

Slides 1-10           Introduction – Simon de Lusignan  PDF

Slides 11-20         Conceptual framework – Siaw-Teng Liaw   PDF

Slides 21-40         Moving from informal to formal methods – teamwork   – Craig Kuziemsky  PDF

Slides 41-54         Formal use of ontologies – Alireza Rahimi  & Gergely Héja  PDF

Slides 55-77         Practical, informal use of an ontological approach in diabetes – Nouf Alharbi &
                            Simon de Lusignan

Slides 78-84         Data quality – do we need ontologies? – Carlos Sáez Silvestre  PDF 

Slide 85-114         Data quality management & governance – including examples – Siaw-Teng Liaw PDF

Slides 115-119     Summary & core group contact details PDF

 3. Audio file of part of the workshop:

An audio file recorded with the verbal consent of the workshop attendees can be downloaded here: 

Ontology-Audio-1     Ontology-Audio-2       Ontology-Audio-3       Ontology-Audio-4

4. Round 1 Task – Identify elements of a toolkit to support the use of ontologies in research or QI

  • Round 1 is intended to be an inclusive process – Round 2 will be selective. Round 2 is the stage at which we will be looking to form a consensus about which tools should remain. 
  • Subsequent to the workshop S-TL and SdeL have suggested the following framework:
  1. The scope of this work is patient centred ontologies in primary and integrated care i.e. the patient is in the centre of our ontological framework 
  2. Though we would eventually like to broaden this work – in this first step we want to restrict our ontologies to those relevant to chronic disease management, and if greater specificity is needed diabetes
  3. We wish to include both informal and formal approaches to using ontologies.  The differences between informal and certified canadian viagra advice formal can be described in two ways:

            Both informal and formal take an ontological world view 

  • Formal methods should be executable, without losing patient preference or autonomy (i.e. pathways and mappings must be patient enabling not prescriptive) 
  • Informal methods should include as a minimum a taxonomy, vocabulary and some form of validation 
  • Generally, the informal approach will be a sub-set of the formal 
  • The toolkit will have five main conceptual areas (for delivering patient-centred integrated care):
  1. To define the fastapps.pl clinical process and decision points
  2. To describe the factors providing continuity care
  3. To deliver holistic careincluding, for chronic disease, the features of the chronic care model
  4. To explore the elements that support or inhibit effective team-working
  5. To describe the data quality pathway from clinical encounter to final curation of data
  • The toolkit will also suggest that the following practical tools/methods/approaches are used 
    • The informal approach should include (1) Conceptualisation – which defines concepts, their relationships, and their attributes; and arranges them into taxonomies; (2) A controlled or reference vocabulary; (3) Subject to some attempt at validation.  Without these three elements we don’t think this can be defined as ontological. 

(1) Conceptualisation

The conceptualisation will include definition of the following:


/ Concepts 

Concepts (either physical/specific or abstract/conceptual).

Concepts i.e. classes


Association between concepts. usually binary.

Relations among the classes

Attributes /Properties


Describe the features of the concepts.

Attributes and which values they can take


Classes are organized into taxonomies, defining superclass- subclass hierarchy

classes organized into a taxonomy, preferably patient-centred

(2) Controlled/reference vocabulary.  For example, SNOMED CT has often been used as a controlled vocabulary.

(3) Validation/critical appraisal/pilot study.  Any ontology must be subject to a crucial appraisal process – to ensure that it meets the

The formal methods should set out steps for: 

 Specification including the vocabulary for the concepts and their relationships

  1. Conceptualisation /competency formulation
  2. Formalisation – Ontologies need to be formalised in a reproducible way, but without losing their patient-centred orientation.
  3. Validation (competency check list): the validation/evaluation is both technical (automated reasoners) and qualitative (with people)

The formal methods should also set out the tools they will use – our core group use the following:  

  • DOLCE – as a high level conceptual tool: DOLCE (Descriptive Ontology for Linguistic and Cognitive Engineering), a descriptive upper-level ontology, is especially designed for automatic reasoning and interoperability.
  • OWL (Web ontology language) http://www.w3.org/TR/owl-ref/

5. Round 2 Task – Consensus about the necessity of each element identified in Round 1. 

  • Round 2 will follow after the completion of Round 1 
  • We will seek to develop a consensus about the appropriateness of each component of the list of components identified in Round 1. 
  • We will use the RAND/UCLA Appropriateness Method (2001)  http://www.rand.org/pubs/monograph_reports/MR1269

Prof Siaw-Teng LIAW University of New South Wales

Prof Simon de LUSIGNAN University of Surrey


OBO Foundry

YBMI 2013 – Supplementary Data File  

 Identify elements of a toolkit to support the use of ontologies in research or QI

–          This will be an inclusive list