Turning Guidelines into Practice: Making It Happen With Standards - Part 2

• Part 2: Solutions For Guideline-based Decision Support
• Improvements to the Guidelines
• Improvements to the Translation Process
• Needs and Benefits of Standards and Architectures
• GEM
• Advances to GEM and the development of advanced authoring tools
• The New Zealand Experience
• Progress to Date and Results Achieved
• References

Click here for Part 1 of this paper

Part 2: Solutions For Guideline-based Decision Support
Solutions to the obstacles facing effective development of guideline-based electronic clinical decision support (ECDS) systems can be considered in two areas: improvement to the guidelines and improvements to the translation process, each of which are considered below.

Improvements to the Guidelines
Several initiatives are underway that have the goal of "improving the guidelines":

The AGREE^{[ a ]} instrument is a 23-item instrument for guideline quality appraisal. The instrument is the product of international effort and is maintained by the Guidelines International Network^{[ b]}.

The GuideLine Implementability Appraisal (GLIA) is an instrument under development by the Yale Guidelines Review Group. It predicts barriers to implementation and is intended for use by developers and implementers. The GLIA is oriented to individual recommendations and is complementary to the AGREE instrument.

The COGS (Conference on Guideline Standardization) checklist is a consolidated list of 18 topics seen as universally necessary components of practice guidelines It is designed for use by developers and identifies items "necessary" for validity and/or usability. As such, this checklist provides a framework to support more comprehensive documentation of practice guidelines.

There are available systematic definitions of recommendation strength, such as GRADE and AAP Policy Statement Classifying Recommendations for Clinical Practice Guidelines, that measure evidence quality. They provide a balance between evidence quality and benefits and risks, costs and harms that are anticipated when the guidelines are implemented.

Improvements to the Translation Process
The need to integrate guidelines seamlessly into clinical workflow and decision support systems has been the driver behind numerous projects worldwide, each using different approaches in guideline representation architecture and implementation.

The absence of commonly agreed standards has created major difficulties for guideline implementers and decision-support systems designers.^[ 7]

Standards are now seen as critical to improvements in the process of translating guidelines for ECDS.

Needs and Benefits of Standards and Architectures
While translation is a critical step, the overall process of decision support development goes well beyond the translation of guidelines, and the implementation of a guideline architecture is of significant benefit to the decision support development process. Implementation of a guideline architecture provides:

• content standards: creation, maintenance and dissemination 
• durability of content: extending beyond the systems which deliver it
• transportability and sharing of information between providers 
• ability to localise content
• interoperability within diverse clinical information systems
• reusability for other related solutions
• efficiency in development and execution.

Importantly, a guideline architecture delivers form and functionality as illustrated in figure 5.

Figure 5: A guideline architecture delivers form and functionality

The transformation of clinical knowledge that is originally expressed in the form of a guideline to a computable format is a major obstacle to the integration of guideline-based knowledge about best practices into computerised clinical decision support systems.^{[ 8 ]}

A variety of representation models that facilitate the computer-based implementation of medical knowledge have been published, including the Guideline Elements Model (GEM) and the Arden Syntax for Medical Logic Modules (MLMs). The GEM representation model is discussed in further detail below.

GEM
GEM provides a methodology for such a transformation. GEM is an XML (eXtensible Markup Language)-based guideline document model that can store and organise the heterogeneous information contained in practice guideline documents. GEM is an international ASTM.^{[ c ]} standard for the representation of practice guidelines in XML format.

GEM is intended to facilitate the translation of natural language guideline documents into a standard computer interpretable format. A strength of the GEM format is that it enables information about the guideline and its development process to be stored in a structured manner and well as encoding guideline recommendations and their supporting evidence. GEM-encoding of guideline knowledge is pursued through a mark-up process that does not require programming knowledge.

GEM Cutter is an XML editor that facilitates guideline mark-up. The software (plus user manual) can be downloaded for evaluation purposes from the GEM web site at Yale University (http://ycmi.med.yale.edu/GEM/), subject to a license agreement.

GEM is intended to be used throughout the entire guideline lifecycle to model information pertaining to guideline development, dissemination, implementation and maintenance. Information at both high and low levels of abstraction can be accommodated. GEM preserves the intent of guideline developers by marking up the actual guideline language.

GEM is constructed as a hierarchy with more than 100 discrete elements and more than nine major branches (including Identity, Developer, Purpose, Intended Audience, Target Population, Method of Development, Testing, Review Plan and Knowledge Components).

GEM is one of an array of tools for improving guideline quality. Guideline translation remains difficult but GEM or systems like GEM may provide part of the solution.

Advances to GEM and the development of advanced authoring tools
A collaboration between the Yale Center for Medical Informatics (YCMI) and Enigma Publishing Limited is developing an authoring tool to improve guideline knowledge elicitation.

The tool will provide an authoring environment for practice guidelines and will incorporate a groupware infrastructure and a rule-authoring wizard which will enforce COGS compliance. It will comply with the GEM guideline architecture.

This work is designed to the improve efficiency and effectiveness of the processes of translation and localisation of formalised guidelines, while informing the further development of the GEM architecture. It will build on the existing body of knowledge related to the use of decision tables to assist these processes while providing the ability to define the values of "critical determinants" and enumerate clinically relevant combinations

The New Zealand Experience
As a country, New Zealand offers certain unique attributes that lend themselves to the evaluation of guideline-based ECDS including:

• a small population (circa four million) and the government as the dominant funder of the health system
• good comprehension of evidence-based medicine at all levels including the Ministry of Health, District Health Boards (DHBs^{[ d ]}) and individual practitioners
• a nationwide commitment to the adoption and use of international informatics standards
• widespread recognition of the value of and promotion of international collaboration 
• a well-established patient identification system in the unique National Health Identifier (NHI) which supports patient movement, patient registries and outcomes measurement
• an internationally leading guidelines organisation (New Zealand Guidelines Group).

New Zealand has been the setting for the development and implementation of a web-based clinical decision support system for cardiovascular disease risk assessment and management for individual high-risk patients in primary care, known as PREDICT^TM-CVD. The solution, which is fully integrated within the electronic patient management system, is the product of a broad collaboration of health players in New Zealand including:

• The University of Auckland
• Counties Manukau DHB (CMDHB Chronic Care Management team
• ProCare Health^{[ e ]}
• New Zealand Guidelines Group
• National Heart Foundation
• New Zealand Ministry of Health
• Enigma Publishing, a private provider of online health knowledge systems
• HealthTech Ltd, a private provider of electronic medical record solutions

PREDICT^TM-CVD offers:

• systematic CVD risk assessment
• evidence-based decision support based on individual inpatient-specific profiles. This includes concise GP recommendation and suggested actions with the ability to access more complex supporting details and references
• personalised, evidence-based, patient advice
• tailored patient information including web links to leaflets, cardioprotective diet details, etc
• collection of non-identifiable patient data
• thepotential for validation or development of new risk equations for whole or sub-populations
• adaptability for the future.

The development and implementation of PREDICT^TM-CVD and its positive impacts are detailed in the paper "Online Management of Cardiovascular Risk in New Zealand with PREDICT^TM– Getting Evidence To the Moment of Care", also in this edition.

Specific challenges that needed to be addressed in PREDICT^TM-CVD development included the management of:

• guideline and evidence issues, including the management of gaps and logic conflicts in the nine guidelines used
• guideline translation 
• an effective, transparent governance process
• an effective clinical quality management process.

The PREDICT^TM platform provided supportive tools and a structure for translating guidelines into specific actions and recommendations for specific patient populations, and for functional delivery of a patient specific care plan, clinical management advice and tailored patient information at the point of care (refer figure 6).

Figure 6: Component suites of the PREDICT^TM solution

Notes:

• PREDICT^TM authoring suite for translation of guideline/best practice into ECDS rules and content, including authoring support tools and a form builder for creating and testing a user template.
• PREDICT^TM engine – links clinical data input to the engine with structured guideline knowledge stored within the systems databases to deliver a patient specific care plan and supporting management information.
• PREDICT^TM quality management suite to ensure quality assurance of clinical content and technical function and to manage version control.
• PREDICT^TM standards management suite to ensure maintenance of accepted technical standards in the PREDICT^TM applications.
• PREDICT^TM localisation tools and processes to support localisation of an approved national PREDICT^TM module.

Progress to Date and Results Achieved
ECDS development and implementation work over the last three to four years, using PREDICT^TM-CVD and supported by the New Zealand Ministry of Health, has already delivered ECDS solutions with evidence of the following system capabilities:

• direct impact on key Ministry of Health strategies in primary care, public health, referrals, Maori health and chronic disease
• assistance in the implementation of DHB/PHO strategies for population health management
• a clinical decision support system which is:
• integral to clinical workflow
• generalisable and scalable 
• applicable to multiple diseases (CVD, diabetes, gynaecology, etc)
• applicable to multiple clinical settings (primary, secondary, community)
• discreet, defined and documented (ie, it truly is a platform for generalisable ECDS)
• effective performance in clinical practice (ProCare, CMDHB, referrals)
• effective generation of stakeholder support (ProCare, CMDHB, the Rotorua General Practice group, the HealthWest PHO, etc)
• improved data collation.

ECDS development in New Zealand has succeeded because of a shared, documented strategic vision, an agreed definition of clinical decision support and its strategic value, widescale sector commitment and the development and implementation of a strategic plan outlining:

• key systems components
• capability requirements 
• infrastructure requirements
• roles and responsibilities and their allocation
• co-ordinated management of development, delivery, implementation, management and evaluation
• project economics: aligned clinical / financial incentives to speed adoption and drive positive behaviour change + ensure sustainability
• standards, process and governance
• communication processes and processes for sector involvement.

A key finding from the project relates to the requirement for and role of standards in ECDS development and implementation.

There has been significant value in the evolving debate on guideline formalisms (for a discussion on guideline formalisms refer to the paper "Guideline Representation Formalism and Electronic Decision Support Systems: Addressing the Guideline - Implementation Gap"  in this edition).

In addition, the process has highlighted the importance of exploring related standards for electronic health records (EHR), clinical document architecture (CDA), messaging and interfacing. Given the inter-relatedness of these different areas of standards development it is critical to gain agreement and begin implementation of guideline standards which will contribute to the stability and conformity required necessary for ongoing advances in ECDS development.

In passing it is worth noting some additional features of ECDS projects which experience has shown make a significant difference to success. ECDS projects should:

• be developed in disease areas with:
• an identified and agreed clinical quality issue
• features which make the area amenable to improvement through the use of ECDS, ie have a need for one or more of the following:
• improved decision making
• reduced error rate
• improved clinical workflow
• improved data capture / monitoring
• have existing agreed guidelines or protocols which lay out the process for improved management
• provide a clear return on investment
• be able to be implemented within the limitations of existing clinical information systems
• have strategic commitment from key decision makers
• be adequately resourced in terms of staff, capital, time and funds.

References

1. McGlynn EA, Asch SM, Adams J, et al. The quality of health care delivered to adults in the United States. N Engl J Med. 2003 Jun 26; 348(26): 2635-45.
2. Grimshaw JM, Russell IT. Effect of clinical guidelines on medical practice: a systematic review of rigorous evaluations. Lancet. 1993 Nov 27; 342(8883): 1317-22.
3. Zielstorff RD. Online practice guidelines: Issues, obstacles, and future prospects. J Am Med Inform Assoc 1998; 5:227–236.
4. Cabana MD, Rand CS, Powe NR, et al. Why don’t physicians follow clinical practice guidelines? A framework for improvement. JAMA. 1999 Oct 20;282(15):1458-65.
5. Shaneyfelt TM, Mayo-Smith MF, Rothwangl J. Are guidelines following guidelines? The methodological quality of clinical practice guidelines in the peer-reviewed medical literature. JAMA. 1999; 281:1900–1905.
6. Sintchenko V, Garsden H. Clinical Decision Support: New Approaches To Usability Study. HIC 2002 Conference, Melbourne, August 2002.
7. Jenders RA, et al. Requirements for and work toward an international standard for representing clinical guidelines. Medinfo 2004. 
8. Gershkovich P, Shiffman RN. An implementation framework for GEM encoded guidelines. Bakken S, ed. Proc AMIA Symp. 2001: 204–8.

Click here for Part 1 of this paper

Footnote