Open Competition 3 - Information Technology
A Distributed Information Architecture for Clinical Practice and Medical Research
Develop a software architecture for physicians and researchers that automatically extracts patient data from electronic medical records - regardless of location, database, or computer code - generates a list of patient problems, and displays information in ways that support diagnostic and therapeutic decisionmaking.
Sponsor: Medaxis Corporation924 Westwood Boulevard
Los Angeles, CA 90024
Quality health care begins with timely, complete information - still a rarity in most hospitals, clinics, and doctors' offices. Proliferation of different - and, usually, incompatible - information systems, along with the growing array of niche technologies and specialty-specific standards, often means that crucial patient data are not available at the point of care. Consequences include redundant examinations, inadequate coordination of services, poor integration of test results, and even missed diagnoses. Medaxis proposes to develop a set of technologies that not only can link and compile widely distributed fragments of electronic medical records but also organize and display the assembled information in ways that support diagnosis and treatment decisions. It intends to develop and demonstrate a distributed information architecture and associated software technologies that will enable physicians and researchers to retrieve pieces of patient information from varied electronic medical records, regardless of database, data type, or definition. The prototype system will consist of four major components. The first will use the extensible markup language (XML) to query and retrieve data from disparate sources. The second will automatically detect and remove patient identifiers in records whenever regulatory standards require confidentiality. The third will analyze the compilation of patient-specific data and then identify and display all relevant health problems, complete with descriptions, findings, interpretations, and links to original reports. The fourth will integrate and visually present patient data in formats that aid pursuit of a specified line of inquiry, such as family history of heart disease or diabetes, for example. Prototype systems will be evaluated at a community-based health care facility dedicated to an employer group and at an academic medical center. Technical risks include achieving an accurate and complete ontology (specified relations among a set of words and phrases for extracting and organizing data) capable of generating high-quality problem lists and constructing algorithms that present data visually in a context that supports decisionmaking. ATP funding will enable Medaxis to assemble the multi-disciplinary team necessary to accomplish its technology goals and to advance the envisioned system to the stage where prospective applications are within the time frames of venture capital firms and other potential investors. Anticipated benefits are many. For example, improved information management could reduce preventable medical errors, reported to cost $29 billion in 1999, and provide researchers with better access to clinical data.