The guideline interchange format: a model for representing guidelines.

OBJECTIVE: To allow exchange of clinical practice guidelines among institutions and computer-based applications. DESIGN: The GuideLine Interchange Format (GLIF) specification consists of GLIF model and the GLIF syntax. The GLIF model is an object-oriented representation that consists of a set of classes for guideline entities, attributes for those classes, and data types for the attribute values. The GLIF syntax specifies the format of the test file that contains the encoding. METHODS: Researchers from the InterMed Collaboratory at Columbia University, Harvard University (Brigham and Women's Hospital and Massachusetts General Hospital), and Stanford University analyzed four existing guideline systems to derive a set of requirements for guideline representation. The GLIF specification is a consensus representation developed through a brainstorming process. Four clinical guidelines were encoded in GLIF to assess its expressivity and to study the variability that occurs when two people from different sites encode the same guideline. RESULTS: The encoders reported that GLIF was adequately expressive. A comparison of the encodings revealed substantial variability. CONCLUSION: GLIF was sufficient to model the guidelines for the four conditions that were examined. GLIF needs improvement in standard representation of medical concepts, criterion logic, temporal information, and uncertainty.

Ontology-based configuration of problem-solving methods and generation of knowledge-acquisition tools: application of PROTEGE-II to protocol-based decision support.

PROTEGE-II is a suite of tools and a methodology for building knowledge-based systems and domain-specific knowledge-acquisition tools. In this paper, we show how PROTEGE-II can be applied to the task of providing protocol-based decision support in the domain of treating HIV-infected patients. To apply PROTEGE-II, (1) we construct a decomposable problem-solving method called episodic skeletal-plan refinement, (2) we build an application ontology that consists of the terms and relations in the domain, and of method-specific distinctions not already captured in the domain terms, and (3) we specify mapping relations that link terms from the application ontology to the domain-independent terms used in the problem-solving method. From the application ontology, we automatically generate a domain-specific knowledge-acquisition tool that is custom-tailored for the application. The knowledge-acquisition tool is used for the creation and maintenance of domain knowledge used by the problem-solving method. The general goal of the PROTEGE-II approach is to produce systems and components that are reusable and easily maintained. This is the rationale for constructing ontologies and problem-solving methods that can be composed from a set of smaller-grained methods and mechanisms. This is also why we tightly couple the knowledge-acquisition tools to the application ontology that specifies the domain terms used in the problem-solving systems. Although our evaluation is still preliminary, for the application task of providing protocol-based decision support, we show that these goals of reusability and easy maintenance can be achieved. We discuss design decisions and the tradeoffs that have to be made in the development of the system.

Participatory design and an eligibility screening tool.

For most medical informatics software products, insufficient effort is spent on the design phase of production. However, poor design often leads to systems that are either not well accepted, or far less effective than they could be. In this paper, we describe the ideas of participatory design and discuss why these ideas are especially applicable to medical informatics systems. In particular, we present a case study in the area of clinical trial protocol management. We designed and developed a tool aimed at increasing accrual to clinical trial protocols at an oncology center. However, the design evolved over time, and features of this design were only discovered through iterative development and interaction with the users within the context of the workplace.

A rational reconstruction of INTERNIST-I using PROTEGE-II.

PROTEGE-II is a methodology and a suite of tools that allow developers to build and maintain knowledge-based systems in a principled manner. We used PROTEGE-II to reconstruct the well-known INTERNIST-I system, demonstrating the role of a domain ontology (a framework for specification of a model of an application area), a reusable problem-solving method, and declarative mapping relations in creating a new, working program. PROTEGE-II generates automatically a domain-specific knowledge-acquisition tool, which, in the case of the INTERNIST-I reconstruction, has much of the functionality of the QMR-KAT knowledge-acquisition tool. This study provides a means to understand better both the PROTEGE-II methodology and the models that underlie INTERNIST-I.

Making generic guidelines site-specific.

Health care providers are more likely to follow a clinical guideline if the guideline's recommendations are consistent with the way in which their organization does its work. Unfortunately, developing guidelines that are specific to an organization is expensive, and limits the ability to share guidelines among different institutions. We describe a methodology that separates the site-independent information of guidelines from site-specific information, and that facilitates the development of site-specific guidelines from generic guidelines. We have used this methodology in a prototype system that assists developers in creating generic guidelines that are sharable across different sites. When combined with site information, generic guidelines can be used to generate site-specific guidelines that are responsive to organizational change and that can be implemented at a level of detail that makes site-specific computer-based workflow management and simulation possible.

Cross-tool communication: from protocol authoring to eligibility determination.

To be effective, informatics tools for clinical trial protocols must inter-operate and share knowledge. We demonstrate a simple XML-based communication of eligibility criteria information between two independently-developed informatics tools. Using a shared DTD model of criteria, an authoring tool (developed within the Protégé environment) can send a list of eligibility criteria to a commercial system for automatic eligibility determination (the "iKnowChart" system by iKnowMed). The criteria model, developed as a Protégé ontology, includes both the terminology and the logic needed to compute eligibility for a given patient. As a demonstration of cross-tool communication, we have encoded criteria from an active clinical trial protocol (E1199), and shown how use of the authoring tool can effectively update the eligibility knowledge and the behavior of the commercial iKnowChart system. As part of the cross-tool knowledge sharing, we use Common Data Elements, an oncology terminology developed by the National Cancer Institute.

A web-based architecture for a medical vocabulary server.

For health care providers to share computing resources and medical application programs across different sites, those applications must share a common medical vocabulary. To construct a common vocabulary, researchers must have an architecture that supports collaborative, networked development. In this paper, we present a web-based server architecture for the collaborative development of a medical vocabulary: a system that provides network services in support of medical applications that need a common, controlled medical terminology. The server supports vocabulary browsing and editing and can respond to direct programmatic queries about vocabulary terms. We have tested the programmatic query-response capability of the vocabulary server with a medical application that determines when patients who have HIV infection may be eligible for certain clinical trials. Our emphasis in this paper is not on the content of the vocabulary, but rather on the communication protocol and the tools that enable collaborative improvement of the vocabulary by any network-connected user.

PROTEGE-II: computer support for development of intelligent systems from libraries of components.

PROTEGE-II is a suite of tools that facilitates the development of intelligent systems. A tool called MAiTRE allows system builders to create and refine abstract models (ontologies) of application domains. A tool called DASH takes as input a modified domain ontology and generates automatically a knowledge-acquisition tool that application specialists can use to enter the detailed content knowledge required to define particular applications. The domain-dependent knowledge entered into the knowledge-acquisition tool is used by assemblies of domain-independent problem-solving methods that provide the computational strategies required to solve particular application tasks. The result is an architecture that offers a divide-and-conquer approach that separates system-building tasks that require skill in domain analysis and modeling from those that require simple entry of content knowledge. At the same time, applications can be constructed from libraries of component--of both domain ontologies and domain-independent problem-solving methods--allowing the reuse of knowledge and facilitating ongoing system maintenance. We have used PROTEGE-II to construct a number of knowledge-based systems, including the reasoning components of T-Helper, which assists physicians in the protocol-based care of patients who have HIV infection.

Tool support for authoring eligibility criteria for cancer trials.

A critical component of authoring new clinical trial protocols is assembling a set of eligibility criteria for patient enrollment. We found that clinical protocols in three different cancer domains can be categorized according to a set of clinical states that describe various clinical scenarios for that domain. Classifying protocols in this manner revealed similarities among the eligibility criteria and permitted some standardization of criteria based on clinical state. We have developed an eligibility criteria authoring tool which uses a standard set of eligibility criteria and a diagram of the clinical states to present the relevant eligibility criteria to the protocol author. We demonstrate our ideas with phase-3 protocols from breast cancer, prostate cancer, and non-small cell lung cancer. Based on measurements of redundancy and percentage coverage of criteria included in our tool, we conclude that our model reduces redundancy in the number of criteria needed to author multiple protocols, and it allows some eligibility criteria to be authored automatically based on the clinical state of interest for a protocol.