Semantic integration of information in a physician's workstation.

Patient care is an information-intensive activity, yet physicians have few tools to effectively access and manage patient data. We studied physicians' information needs in an outpatient clinic, and developed a prototype physician's workstation (PWS) to address those needs. The PWS provides integrated access to patient information and uses embedded domain knowledge to enhance the presentation of clinical information to the physician. All the applications in the PWS share a common patient context, defined by the state of the internal patient model--semantic integration. Relevant data are presented together and higher-order alerts are generated by combining notable events with relevant data from the patient context. Semantic integration allows us to present and to operate on all patient data in a given patient's context, significantly enhancing the effectiveness with which information is presented to the physician.

Implementing a Physician's Workstation using client/server technology and the distributed computing environment.

PWS is a physician's workstation research prototype developed to explore the use of information management tools by physicians in the context of patient care. The original prototype was implemented in a client/server architecture using a broadcast message server. As we expanded the scope of the prototyping activities, we identified the limitations of the broadcast message server in the areas of scalability, security, and interoperability. To address these issues, we reimplemented PWS using the Open Software Foundation's Distributed Computing Environment (DCE). We describe the rationale for using DCE, the migration process, and the benefits achieved. Future work and recommendations are discussed.

Automated identification of relevant patient information in a physician's workstation.

The introduction of computer-based patient records offers an opportunity to improve the ability of physicians to browse the medical record and to monitor patient events. We describe a methodology to identify relevant information in the patient record. This methodology combines a patient-specific physiological model with functions to determine relevance of patient information. The model consists of a qualitative representation of physiological parameters and influences, custom-tailored to a particular patient's problems, medications, and test results. We describe two applications of this model in the context of an integrated physician's workstation: automatic linking of relevant patient information for configuration of user-interface displays, and monitoring of patient events to prevent oversight of noteworthy information.

Heterogenous database integration in a physician workstation.

We discuss the integration of a variety of data and information sources in a Physician Workstation (PWS), focusing on the integration of data from DHCP, the Veteran Administration's Distributed Hospital Computer Program. We designed a logically centralized, object-oriented data-schema, used by end users and applications to explore the data accessible through an object-oriented database using a declarative query language. We emphasize the use of procedural abstraction to transparently integrate a variety of information sources into the data schema.

An open systems architecture for development of a physician's workstation.

We are developing a physician's workstation consisting of highly integrated information management tools for use by physicians in patient care. We have designed and implemented an open systems, client/server architecture as a development platform which allows new applications to be easily added to the system. Applications cooperate by exchanging messages via a broadcast message server.

Physicians' workstations: integrated information management for clinicians.

This paper describes a project to specify, design, develop, and evaluate a physician's workstation for use in patient care. We conducted an ethnographic study of physicians' information needs in an outpatient setting, from which we derived a set of functional specifications for a physician's workstation. We have implemented an experimental prototype using an open systems, client/server architecture, and are exploring research issues in heterogeneous database integration, object-oriented database technology, model-based reasoning, and semantic integration. We plan to evaluate our workstation prototype in a clinical setting to assess its impact on quality of care and health care costs.