Modelling clinical narrative as computable knowledge: The NICE computable implementation guidance project

Area of Science:

• Health Informatics
• Clinical Knowledge Management
• Digital Health Implementation

Background:

• Translating narrative clinical guidelines into computable knowledge presents a significant challenge.
• The UK National Institute for Health and Care Excellence (NICE) Content Advisory Board (CAB) aims to enhance clinical decision support, audit, and guideline development.
• A transition towards natively digital content production by NICE was explored.

Purpose of the Study:

• To explore technical approaches for transitioning NICE recommendations into computable knowledge.
• To establish a framework for developing traceable, consistent, and high-fidelity digital clinical guidance.
• To guide software developers and support local practice audits.

Main Methods:

• The NICE Computable Implementation Guidance (NCIG) project was established following a 'collaborathon'.
• Workstreams focused on user stories, information modeling, and output formats.
• The WHO Digital Adaptation Kit (DAK) was adopted as a technology-agnostic method for logical modeling.

Main Results:

• An intermediate logical model, using the WHO DAK, proved more achievable than direct technical standards.
• The DAK was adapted to model user scenarios, processes, data elements, and decision logic.
• Future work will address indicators and implementation in primary care systems.

Conclusions:

• The WHO DAK, with modifications, is a promising approach for creating technology-neutral logical specifications of NICE recommendations.
• Implementing concurrent computable modeling involves complex but tractable methodological and cultural considerations.
• Successful implementation requires strong leadership and commitment.