Article type
Abstract
Background
The strategy of the UK National Institute for Health and Care Excellence (NICE) for 2021-2026 set out objectives for new methods of guideline development that move toward “recommendations produced in an interactive, digitalised format.” To support these objectives, NICE formed a Content Advisory Board (CAB) comprising a panel of external advisors from industry and academia.
In the existing global IT market, practice guidelines are implemented in clinical decision support (CDS) software by each supplier separately repeating the reverse engineering of guidelines into computable logic. This is highly inefficient and has numerous risks of misinterpretation, inconsistency, interpolation, and poor usability.
Following a series of exploratory “collaborathon” events based on the World Health Organization (WHO) Digital Adaptation Kit (DAK) and learning from the global movement on Mobilizing Computable Biomedical Knowledge (MCBK), NICE has decided to change its guideline development process so that computable logic is addressed prospectively.
Objectives
1. To make NICE recommendations usable in CDS
2. To move from guideline used as reference or learning to guiding point-of-care interactions
3. To define the clinical codes for data collection, enabling reporting back to NICE, helping to constantly learn from implementation to understand impact as part of a Learning Health System
4. To benefit patients not only through improved guideline availability in CDS but also in supporting reliable implementation of guidelines in patient-facing apps
Methods
NICE is implementing this new approach from April 2024. The presentation will report how NICE has built “computability” into the template for all guideline production, modified its governance structures, and implemented education for staff and committees.
Results
This presentation will report progress to date, including several examples of where guideline recommendations have been written differently, how templates and taxonomies have been developed, and feedback from NICE committees about how computability feeds into decision-making. We will also report on how this has started to change the use of NICE guidelines in practice.
Conclusions
This approach also has applicability in global expert domains outside health care and extends the FAIR principles (Findable, Accessible, Interoperable, and Reusable) from data to computable knowledge.
The strategy of the UK National Institute for Health and Care Excellence (NICE) for 2021-2026 set out objectives for new methods of guideline development that move toward “recommendations produced in an interactive, digitalised format.” To support these objectives, NICE formed a Content Advisory Board (CAB) comprising a panel of external advisors from industry and academia.
In the existing global IT market, practice guidelines are implemented in clinical decision support (CDS) software by each supplier separately repeating the reverse engineering of guidelines into computable logic. This is highly inefficient and has numerous risks of misinterpretation, inconsistency, interpolation, and poor usability.
Following a series of exploratory “collaborathon” events based on the World Health Organization (WHO) Digital Adaptation Kit (DAK) and learning from the global movement on Mobilizing Computable Biomedical Knowledge (MCBK), NICE has decided to change its guideline development process so that computable logic is addressed prospectively.
Objectives
1. To make NICE recommendations usable in CDS
2. To move from guideline used as reference or learning to guiding point-of-care interactions
3. To define the clinical codes for data collection, enabling reporting back to NICE, helping to constantly learn from implementation to understand impact as part of a Learning Health System
4. To benefit patients not only through improved guideline availability in CDS but also in supporting reliable implementation of guidelines in patient-facing apps
Methods
NICE is implementing this new approach from April 2024. The presentation will report how NICE has built “computability” into the template for all guideline production, modified its governance structures, and implemented education for staff and committees.
Results
This presentation will report progress to date, including several examples of where guideline recommendations have been written differently, how templates and taxonomies have been developed, and feedback from NICE committees about how computability feeds into decision-making. We will also report on how this has started to change the use of NICE guidelines in practice.
Conclusions
This approach also has applicability in global expert domains outside health care and extends the FAIR principles (Findable, Accessible, Interoperable, and Reusable) from data to computable knowledge.