Experience with an electronic health record for a homeless population.

A computerized electronic medical record (EMR) system using client-server architecture was designed and implemented by the Laboratory of Computer Science for use by the Boston Health Care for the Homeless Program (BHCHP) to meet the unique medical record needs of the homeless. For the past three years, this EMR has been used to assist providers in the delivery of health care to the homeless population of Boston. As the BHCHP has grown and technology improved, it is important to review what features of the EMR work, and to investigate what improvements can be made for the better delivery of care to the homeless, especially as we approach the next century.

Optimizing healthcare research data warehouse design through past COSTAR query analysis.

Over the past two years we have reviewed and implemented the specifications for a large relational database (a data warehouse) to find research cohorts from data similar to that contained within the clinical COSTAR database at the Massachusetts General Hospital. A review of 16 years of COSTAR research queries was conducted to determine the most common search strategies. These search strategies are relevant to the general research community, because they use the Medical Query Language (MQL) developed for the COSTAR M database which is extremely flexible (much more so than SQL) and allows searches by coded fields, text reports, and laboratory values in a completely ad hoc fashion. By reviewing these search strategies, we were able to obtain user specifications for a research oriented healthcare data warehouse that could support 90% of the queries. The data warehouse was implemented in a relational database using the star schema, allowing for highly optimized analytical processing. This allowed queries that performed slowly in the M database to be performed very rapidly in the relational database. It also allowed the data warehouse to scale effectively.

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.

A study of collaboration among medical informatics research laboratories.

The InterMed Collaboratory involves five medical institutions (Stanford University, Columbia University, Brigham and Women's Hospital, Massachusetts General Hospital, and McGill University) whose mandate has been to join in the development of shared infrastructural software, tools, and system components that will facilitate and support the development of diverse, institution-specific applications. Collaboration among geographically distributed organizations with different goals and cultures provides significant challenges. One experimental question, underlying all that InterMed has set out to achieve, is whether modern communication technologies can effectively bridge such cultural and geographical gaps, allowing the development of shared visions and cooperative activities so that the end results are greater than any one group could have accomplished on its own. In this paper we summarize the InterMed philosophy and mission, describe our progress over 3 years of collaborative activities, and present study results regarding the nature of the evolving collaborative processes, the perceptions of the participants regarding those processes, and the role that telephone conference calls have played in furthering project goals. Both informal introspection and more formal evaluative work, in which project participants became subjects of study by our evaluation experts from McGill, helped to shift our activities from relatively unfocused to more focused efforts while allowing us to understand the facilitating roles that communications technologies could play in our activities. Our experience and study results suggest that occasional face-to-face meetings are crucial precursors to the effective use of distance communications technologies; that conference calls play an important role in both task-related activities and executive (project management) activities, especially when clarifications are required; and that collaborative productivity is highly dependent upon the gradual development of a shared commitment to a well-defined task that leverages the varying expertise of both local and distant colleagues in the creation of tools of broad utility across the participating sites.

The Unified Medical Language System: an informatics research collaboration.

In 1986, the National Library of Medicine (NLM) assembled a large multidisciplinary, multisite team to work on the Unified Medical Language System (UMLS), a collaborative research project aimed at reducing fundamental barriers to the application of computers to medicine. Beyond its tangible products, the UMLS Knowledge Sources, and its influence on the field of informatics, the UMLS project is an interesting case study in collaborative research and development. It illustrates the strengths and challenges of substantive collaboration among widely distributed research groups. Over the past decade, advances in computing and communications have minimized the technical difficulties associated with UMLS collaboration and also facilitated the development, dissemination, and use of the UMLS Knowledge Sources. The spread of the World Wide Web has increased the visibility of the information access problems caused by multiple vocabularies and many information sources which are the focus of UMLS work. The time is propitious for building on UMLS accomplishments and making more progress on the informatics research issues first highlighted by the UMLS project more than 10 years ago.

Using web technology and Java mobile software agents to manage outside referrals.

A prototype, web-based referral application was created with the objective of providing outside primary care providers (PCP's) the means to refer patients to the Massachusetts General Hospital and the Brigham and Women's Hospital. The application was designed to achieve the two primary objectives of providing the consultant with enough data to make decisions even at the initial visit, and providing the PCP with a prompt response from the consultant. The system uses a web browser/server to initiate the referral and Java mobile software agents to support the workflow of the referral. This combination provides a light client implementation that can run on a wide variety of hardware and software platforms found in the office of the PCP. The implementation can guarantee a high degree of security for the computer of the PCP. Agents can be adapted to support the wide variety of data types that may be used in referral transactions, including reports with complex presentation needs and scanned (faxed) images Agents can be delivered to the PCP as running applications that can perform ongoing queries and alerts at the office of the PCP. Finally, the agent architecture is designed to scale in a natural and seamless manner for unforeseen future needs.

Component architecture for web based EMR applications.

The World Wide Web provides the means for the collation and display of disseminated clinical information of use to the healthcare provider. However, the heterogeneous nature of clinical data storage and formats makes it very difficult for the physician to use one consistent client application to view and manipulate information. Similarly, developers are faced with a multitude of possibilities when creating interfaces for their users. A single patients records may be distributed over a number of different record keeping systems, and/or a physician may see patients whose individual records are stored at different sites. Our goal is to provide the healthcare worker with a consistent application interface independent of the parent database and at the same time allow developers the opportunity to customize the GUI in a well controlled, stable application environment.

Long-term changes in compliance with clinical guidelines through computer-based reminders.

We evaluate the effectiveness of computer-based reminders in improving compliance with preventive medicine screening guidelines and examine the long-term impact of these reminders. Physicians in an ambulatory care practice were given a summary health maintenance report of preventive screening items at each scheduled patient visit. The electronic medical record (COSTAR) in use in this practice was programmed to integrate 13 clinical guidelines into the routine flow of care. Mean performance of 10 out of 13 health maintenance measures improved in the year following introduction of the integrated guideline report (p less than .001 by chi-square test for 8 items, p less than .01 for 2 items). Five years after the report was introduced, improvement in mean performance persisted for 7 measures (p less than .001 by chi-square test), compliance improved for one additional measure (p less than .001), and improvement disappeared for three measures that had shown improvement in the first year of the intervention.

DXplain on the Internet.

DXplain, a computer-based medical education, reference and decision support system has been used by thousands of physicians and medical students on stand-alone systems and over communications networks. For the past two years, we have made DXplain available over the Internet in order to provide DXplain's knowledge and analytical capabilities as a resource to other applications within Massachusetts General Hospital (MGH) and at outside institutions. We describe and provide the user experience with two different protocols through which users can access DXplain through the World Wide Web (WWW). The first allows the user to have direct interaction with all the functionality of DXplain where the MGH server controls the interaction and the mode of presentation. In the second mode, the MGH server provides the DXplain functionality as a series of services, which can be called independently by the user application program.

An XML portable chart format.

The clinical chart remains the fundamental record of outpatient clinical care. As this information migrates to electronic form, there is an opportunity to create standard formats for transmitting these charts. This paper describes work toward a Portable Chart Format (PCF) that can represent the relevant aspects of an outpatient chart. The main goal of the format is to provide a packaging medium for outpatient clinical charts in a transfer of care scenario. A secondary goal is to support the aggregation of comparable clinical data for outcomes analysis. The syntax used for PCF is Extended Markup Language (XML), a W3C standard. The structure of the PCF is based on a clinically relevant view of the data. The data definitions and nomenclature used are based primarily on existing clinical standards.

Maintaining the confidentiality of medical records shared over the Internet and the World Wide Web.

The Boston Electronic Medical Record Collaborative is working to develop a system that will use the World Wide Web to transfer computer-based patient information to clinicians in emergency departments. Maintaining adequate confidentiality of these records while still facilitating patient care is paramount to this effort. This paper describes an explicit protocol that would make it possible to electronically identify patients and providers, secure permission for release of records, and track information that is transmitted. It is hoped that other, similar efforts now underway will be able to use and build on this model. Comment on this proposal is invited from all parties with an interest in confidentiality. The system will be used only with "scrubbed" data-data from which all identifiers have been removed-until it is generally agreed that the confidentiality methods proposed here are appropriate and sufficient.

"Just-in-time" clinical information.

The just-in-time (JIT) model originated in the manufacturing industry as a way to manage parts inventories process so that specific components could be made available at the appropriate times (that is, "just in time"). This JIT model can be applied to the management of clinical information inventories, so that clinicians can have more immediate access to the most current and relevant information at the time they most need it--when making clinical care decisions. The authors discuss traditional modes of managing clinical information, and then describe how a new, JIT model may be developed and implemented. They describe three modes of clinician-information interactions that a JIT model might employ, the scope of information that may be made available in a JIT model (global information or local, case-specific information), and the challenges posed by the implementation of such an information-access model. Finally, they discuss how JIT information access may change how physicians practice medicine, various ways JIT information may be delivered, and concerns about the trustworthiness of electronically published and accessed information resources.

Using software agents to maintain autonomous patient registries for clinical research.

A software agent is an application that can function in an autonomous and intelligent fashion. We have used mobile software agents to maintain clinicians' patient research databases (patient registries). Agents were used to acquire data from the clinician and place it into the registries, copy data from hospital databases into the registries, and report data from the registries. The agents were programmed with the intelligence to navigate through complex network security, interact with legacy systems, and protect themselves from various forms of failure at multiple levels. To maximize the separation between our system and the hospital information infrastructure we often used Java, a platform-independent language, to program and distribute our software agents. By using mobile agents, we were able to distribute the computing time required by these applications to underutilized host machines upon which the registries could be maintained.

Evaluation of a decision support system for pressure ulcer prevention and management: preliminary findings.

A decision support system for prevention and management of pressure ulcers was developed based on AHCPR guidelines and other sources. The system was implemented for 21 weeks on a 20-bed clinical care unit. Fifteen nurses on that unit volunteered as subjects of the intervention to see whether use of the system would have a positive effect on their knowledge about pressure ulcers and on their decision-making skills related to this topic. A similar care unit was used as a control. In addition, the system was evaluated by experts for its instructional adequacy, and by end users for their satisfaction with the system. Preliminary results show no effect on knowledge about pressure ulcers and no effect on clinical decision making skills. The system was rated positively for instructional adequacy, and positively for user satisfaction. User interviews related to satisfaction supplemented the quantitative findings. A discussion of the issues of conducting experiments like this in today's clinical environment is included.

MedWeaver: integrating decision support, literature searching, and Web exploration using the UMLS Metathesaurus.

Integrating functions from disparate and widely-distributed information systems has been an interest of the medical informatics community for some time. Barriers to progress have included the lack of network-accessible information sources, inadequate methods for inter-system messaging, and lack of vocabulary translation services. With the advent of the World Wide Web (WWW) and the evolution of the National Library of Medicine's Unified Medical Language System (UMLS), it is now possible to develop applications that integrate functions from diverse, distributed systems. In this paper we describe one such system, MedWeaver, a WWW application that integrates functions from a decision support application (DXplain), a literature searching system (WebMedline), and a clinical Web searching system (CliniWeb) using the UMLS Metathesaurus for vocabulary translation. This system demonstrates how application developers can design systems around anticipated clinical information needs and then draw together the needed content and functionality from diverse sources.

Puya: a method of attracting attention to relevant physical findings.

Puya is a method that compares the physical exam in an electronic clinical note with a set of stereotypical physical exam sentences that have been previously classified as "normal". The note is then displayed in a web browser with normal findings clearly delineated. The list of stereotypical sentences comes from a set of physical findings found within extensive electronic medical record. This list is then screened to select only those that represent "normal" findings, a process that yields 96% total agreement among 4 clinicians surveyed. This final list of stereotypical "normal" sentences accounts for 64% of the clinical narrative text. Sentences in the clinical note that do not match sentences in the "normal" list are assumed to be "abnormal". Puya screened 98 clinical notes consisting of 610 individual sentences. Puya achieved a sensitivity of 100%, a specificity of 63%, a positive predictive value of 44% and a negative predictive value of 100%. This leads to an application that reduces informational noise.

Maintaining the integrity of a Web-based medical vocabulary glossary.

Medical vocabulary databases are vital components of electronic medical record (EMR) systems. While their performance and efficiency has been extensively explored by many authors, few have dealt with the maintenance of their semantic integrity. Clinicians' preference for an optimistic system has introduced a need for monitoring and filtering proposed additions to the vocabulary tables. We propose the use of batch processing and memo-posting as means of implementing a World-Wide-Web-based Controlled Vocabulary Glossary as a monitored optimistic system.

Design considerations in migrating an obstetrics clinical record to the Web.

Recently the American College of Obstetricians and Gynecologists (ACOG) embarked on an effort to promote the development of nationally networked obstetrical records. The Laboratory of Computer Science (LCS) is collaborating with them to help achieve this goal through the development of a web-based prototype of an electronic medical record (EMR) which would allow the entry and display of typical clinical information for the obstetric patient. The process of porting a stand alone application to the web environment necessitated the development of a robust software scheme that could exploit the strengths of Web-based technologies and avoid some of the drawbacks inherent in a stateless environment.

A knowledge-based decision support system for prevention and treatment of pressure ulcers.

As part of a research project intended to provide problem-based knowledge to clinicians at the point of care, we have developed a system that supports the nurse's development of patient-specific, guideline-based treatment plans for patients who have pressure ulcers or are at risk for developing them. The system captures coded data about assessment, diagnosis and interventions using a point-and-click interface. Knowledge is accessible to the user via: 1) hypertext links from the data entry screens; 2) explicit entry into an indexed version of the guideline; 3) imbedded knowledge-based rules that critique the diagnosis and offer guidance for treatment; and 4) explicit entry into interactive algorithms. The system has been implemented experimentally on one care unit at our hospital, where its impact will be assessed in comparison with a control unit. Data on 113 patients were entered during the 21-week experimental period. The system is being evaluated for its instructional adequacy, its impact on clinicians' decision-making and knowledge, and on processes of care. Users' perceptions of the system are also being evaluated. Dissemination issues in the context of today's health care environment are addressed.