NICE process and methods

5 Developing a conceptual model

5 Developing a conceptual model

A conceptual model is defined by Tappenden as 'the abstraction and representation of complex phenomena of interest in some readily expressible form, such that the individual stakeholders' understanding of the parts of the system, and the mathematical representation of that system, can be shared, questioned, tested and ultimately agreed'[1]. Two different, but interlinked, conceptual model forms can be used: problem-oriented and design-oriented.

It is recommended that each review question should have a linked conceptual model. The 'problem-oriented conceptual' model is a simplified, diagrammatical representation of the care/service pathway that describes the resources, processes and interactions in the delivery of healthcare interventions. The main aims of using a problem-oriented conceptual model are to assist in the understanding of a service in order to identify areas for improvement and to describe existing or planned services to allow a shared understanding of them.

The use of a problem‑oriented conceptual model is a method for exploring the interrelationship between processes and structures, and to ensure that key areas are approached in the most logical and efficient way. It also should help focus questions and reviews to address key areas, as well as providing a useful tool to assess how the various discrete questions are interlinked and how much of the service will be covered. The problem‑oriented conceptual model should not be limited by what is feasible. This problem‑oriented conceptual model can be developed as part of the scoping process (for example, at the stakeholder scoping workshop.

The problem‑oriented conceptual model should be able to contextualise and describe the service in terms of the following areas:

  • who is using the service

  • interventions being delivered

  • current service models being used

  • regional and/or national variations

  • key decision makers

  • key outcomes for the service

  • assumed strengths of the service

  • assumed weaknesses of the service

  • data identification

  • potential trade-offs between options such as effectiveness, volume and impact on travelling times for patients

  • waiting list issues.

In addition, it may also be helpful to consider the disease process being addressed by the service under consideration as part of the problem-oriented conceptual model.

The problem-oriented conceptual model links to a 'design-oriented conceptual model' that is used as the quantitative basis to inform the structure, assumptions and data needs of the computer models ('implementation models') to be used to assess effectiveness and cost‑effectiveness. The design-oriented conceptual model is an explicit simplification and abstraction of the problem-oriented conceptual model, mediated by what is feasible and by the availability of evidence and data. This design-oriented conceptual model can also be used to help structure the review questions and to specify precisely the data and evidence that will be needed by the implementation model to simulate the service decision problem, and so generate appropriate effectiveness and cost‑effectiveness outcomes to inform recommendations.

[1] Tappenden P. Conceptual modelling for health economic model development. ScHARR Discussion Paper (number 12/05) 2012, University of Sheffield.