Salvaging legacy data: mapping an obsolete medical nomenclature to a modern one.

The Veterinary Medical Database (VMDB) is a repository containing abstracts of over six million case records from 24 veterinary colleges throughout the U.S. and Canada. These case record abstracts, spanning almost 40 years, represent a valuable resource for outcomes analysis and hypothesis generation. Database records are currently encoded using the Standard Nomenclature of Veterinary Diseases and Operations (SNVDO), a precoordinated, hierarchical coding system. SNVDO has not been updated since 1977 and is outdated and inadequate to express the current state of medical knowledge. We undertook to manually map a subset of the SNVDO codes to a modern medical nomenclature, SNOMED-RT (Version 1.0), and to evaluate the quality of the resultant mappings and the acceptability of the mapping method used. We found that the distribution of frequency of use of the SNVDO codes in the VMDB records is highly skewed, with a small number of codes accounting for a large percentage of the records. We targeted our mapping efforts on that subset of codes. We found that our targeted manual mapping of the SNVDO codes to SNOMED-RT codes was feasible and produced good quality results, based on separate evaluations performed by two domain experts. However, a significant proportion of the SNVDO codes could not be mapped to a single SNOMED-RT concept, necessitating construction of multiple-code post-coordinated terms. Additionally, this manual mapping was very labor-intensive.

Evaluation of controlled vocabulary resources for development of a consumer entry vocabulary for diabetes.

BACKGROUND: Digital information technology can facilitate informed decision making by individuals regarding their personal health care. The digital divide separates those who do and those who do not have access to or otherwise make use of digital information. To close the digital divide, health care communications research must address a fundamental issue, the consumer vocabulary problem: consumers of health care, at least those who are laypersons, are not always familiar with the professional vocabulary and concepts used by providers of health care and by providers of health care information, and, conversely, health care and health care information providers are not always familiar with the vocabulary and concepts used by consumers. One way to address this problem is to develop a consumer entry vocabulary for health care communications. OBJECTIVES: To evaluate the potential of controlled vocabulary resources for supporting the development of consumer entry vocabulary for diabetes. METHODS: We used folk medical terms from the Dictionary of American Regional English project to create extended versions of 3 controlled vocabulary resources: the Unified Medical Language System Metathesaurus, the Eurodicautom of the European Commission's Translation Service, and the European Commission Glossary of popular and technical medical terms. We extracted consumer terms from consumer-authored materials, and physician terms from physician-authored materials. We used our extended versions of the vocabulary resources to link diabetes-related terms used by health care consumers to synonymous, nearly-synonymous, or closely-related terms used by family physicians. We also examined whether retrieval of diabetes-related World Wide Web information sites maintained by nonprofit health care professional organizations, academic organizations, or governmental organizations can be improved by substituting a physician term for its related consumer term in the query. RESULTS: The Dictionary of American Regional English extension of the Metathesaurus provided coverage, either direct or indirect, of approximately 23% of the natural language consumer-term-physician-term pairs. The Dictionary of American Regional English extension of the Eurodicautom provided coverage for 16% of the term pairs. Both the Metathesaurus and the Eurodicautom indirectly related more terms than they directly related. A high percentage of covered term pairs, with more indirectly covered pairs than directly covered pairs, might be one way to make the most out of expensive controlled vocabulary resources. We compared retrieval of diabetes-related Web information sites using the physician terms to retrieval using related consumer terms We based the comparison on retrieval of sites maintained by non-profit healthcare professional organizations, academic organizations, or governmental organizations. The number of such sites in the first 20 Results from a search was increased by substituting a physician term for its related consumer term in the query. This suggests that the Dictionary of American Regional English extensions of the Metathesaurus and Eurodicautom may be used to provide useful links from natural language consumer terms to natural language physician terms. CONCLUSIONS: The Dictionary of American Regional English extensions of the Metathesaurus and Eurodicautom should be investigated further for support of consumer entry vocabulary for diabetes.

Consumer reports and health care information: a call for research.

The consumer movement together with the information explosion has a tremendous impact upon all aspects of American life. Health care is a prime example of a vital area of importance that has seen and will continue to see a demand for more timely and appropriate information that consumers can use to assist in decision making about personal health care services. This article proposes that cancer, in particular breast cancer, presents one of the major diagnostic areas that has caused women to seek out such information. We propose needed research including a conceptual framework that may guide that research.

Comparing frequency of content-bearing words in abstracts and texts in articles from four medical journals: an exploratory study.

BACKGROUND: Retrieval tests have assumed that the abstract is a true surrogate of the entire text. However, the frequency of terms in abstracts has never been compared to that of the articles they represent. Even though many sources are now available in full-text, many still rely on the abstract for retrieval. METHODS: 1,138 articles with their abstracts were downloaded from Journal of the American Medical Association, New England Journal of Medicine, Lancet, and the British Medical Journal. Words were extracted from the articles and their abstracts and the frequency of each word was counted in both sources. Each article and its abstract were tested using a chi-squared test to determine if the words in the abstract occurred as frequently as would be expected. RESULTS: 96% of the abstracts tested as samples of the article they represented. CONCLUSION: In these four journals, the abstracts are lexical, as well as intellectual, surrogates for the documents they represent.

The natural history of the use of healthcare information by women with breast cancer: a conceptual model.

The overall goal of our research agenda is to contribute to improved quality of healthcare by identifying factors that foster or inhibit the use of healthcare information by patients to make informed healthcare decisions. We propose to study the natural history of the use of healthcare information by women with breast cancer to support decisions about health care. To do so in this paper we propose a conceptual model developed based on an extensive literature review and critique that describes patients' health information use over the disease course. It will guide our further investigation of the complex relationships among patients' personal circumstances, the progress of their medical treatment, and their satisfaction and empowerment as informed decision-makers. The model will help policy makers and health professionals identify the best means to provide patients with useful information, and help all stakeholders in health care acquire information needed to improve healthcare quality.

Comparing frequency of word occurrences in abstracts and texts using two stop word lists.

Retrieval tests have assumed that the abstract is a true surrogate of the entire text. However, the frequency of terms in abstracts has never been compared to that of the articles they represent. Even though many sources are now available in full-text, many still rely on the abstract for retrieval. 1,138 articles with their abstracts were downloaded from Journal of the American Medical Association, New England Journal of Medicine, the British Medical Journal, and the Lancet. Based on two stop word lists, one long and one short, content bearing words were extracted from the articles and their abstracts and the frequency of each word was counted in both sources. Each article and its abstract were tested using a chi-squared test to determine if the words in the abstract occurred as frequently as would be expected. 96% to 98% of the abstracts tested were not significantly different than random samples of the articles they represented. In these four journals, the abstracts are lexical, as well as intellectual, surrogates for the articles they represent.

A World Wide Web-based health resource. Survey of Missouri school nurses to determine priority health information resources for SchoolhealthLink.

Two hundred ninety-two school nurses in Missouri participated in a mailed survey to aid in the design of SchoolhealthLink, a World Wide Web (WWW)-based information service for Missouri school nurses and children. The nurses identified specific health information resources likely to benefit school nurses and school children and prioritized these resources. The school nurses assigned high priority to 11 types of health information resources: individualized health care plans, emergency care plans, communicable disease control plans, acute illness, injuries, communicable diseases, hotline numbers, medications in schools, immunization protocols and standards, community-based health care resources, and Department of Elementary and Secondary Education regulations that affect school nurse practice. The four most common health problems school nurses identified were asthma, attention deficit disorder, diabetes, and head lice. SchoolhealthLink will provide a one-stop WWW-based resource for school nurses and school children.

Development and validation of an exercise performance support system for people with lower extremity impairment.

OBJECTIVE: To identify innovative strategies to support appropriate, self-directed exercise that increase physical activity levels of people with arthritis. This article reports on one interactive, multimedia exercise performance support system (PSS) for people with lower extremity impairments in strength or flexibility. METHODS: An interdisciplinary team developed the PSS using self-report of lower extremity musculoskeletal impairments (flexibility and strength) to produce an individualized exercise program with video and print educational materials. Initial evaluation has investigated the validity and reliability of program assessments and recommendations. RESULTS: PSS self-report and professional assessments were similar, with more impairments indicated by self-report. PSS exercise recommendations were similar to those made by 3 expert physical therapists using the same exercise data base. Results of PSS impairment assessments were stable over a 1-week period. CONCLUSION: PSS exercise recommendations appear to be reliable and a valid reflection of current exercise knowledge in rheumatology. Furthermore, users were able to complete the computer-based program with minimal assistance and reported it to be enjoyable and informative.

Do electronic mail discussion lists act as virtual colleagues?

Anesthesiology Discussion Group (ADG), an electronic mail (email) discussion list, has previously been shown to be a clinically oriented, cost-effective form of telemedicine. ADG is composed of an international collection of anesthesia providers. Discussions with colleagues are generally informal in nature and are examples of types of information-seeking behavior which frequently occur in hallways or lounges of a hospital or clinic. Information-seeking occurs when a health care provider searches for information which will be used to solve or satisfy a patient's problem or need. We surveyed practitioners who had previously submitted non-rhetorical, clinical questions to the group. After analysis of the questionnaire results, we conclude that ADG is a valuable resource used for information-seeking and is a clinically effective form of telemedicine. Many of the respondents indicated that they used ADG to obtain second opinions from the collective expertise of group members. Respondents also indicated that they were generally satisfied with the quality of responses and would not hesitate to use ADG for future clinical questions.

Virtual shelves. II: A unified catalog for a heterogeneous collection.

We describe a unified catalog of traditional and digital resources in medical informatics, using standard cataloging principles (AACR2), schemes (LCC), and coding formats (USMARC). The unified catalog integrates the bibliographic records of physical items in the heterogeneous collection with the bibliographic records of network accessible digital items, using prescribed cataloging formats to new effect. The unified catalog is collections-based. We do not use the MARC 856 field to specify the network location of a digital item. The location of a digital item is determined by mapping its call number through a location guide to a network address. This mapping is strictly isomorphic to the way the shelf location of a physical item is determined within the bounds of a controlled collection.

Delivery of laboratory data with World Wide Web technology.

We have developed an experimental World Wide Web (WWW) based system to deliver laboratory results to clinicians in our Veterinary Medical Teaching Hospital. Laboratory results are generated by the clinical pathology section of our Veterinary Medical Diagnostic Laboratory and stored in a legacy information system. This system does not interface directly to the hospital information system, and it cannot be accessed directly by clinicians. Our "meta" system first parses routine print reports and then instantiates the data into a modern, open-architecture relational database using a data model constructed with currently accepted international standards for data representation and communication. The system does not affect either of the existing legacy systems. Location-independent delivery of patient data is via a secure WWW based system which maximizes usability and allows "value-added" graphic representations. The data can be viewed with any web browser. Future extensibility and intra- and inter-institutional compatibility served as key design criteria. The system is in the process of being evaluated using accepted methods of assessment of information technologies.

Using concept maps on the World-Wide Web to access a curriculum database for problem-based learning.

Development of medical school curriculum databases continues to be challenging. Representation of the instructional unit is becoming increasingly difficult due to characteristics of the problem-based learning (PBL) curricula. Curriculum databases may be used to store materials for the PBL curricula, and also to provide a delivery mechanism for those materials. However, in order to take advantage of the curriculum database as a tool for PBL, methods for accessing the curriculum database that are better suited to the information needs of students, faculty, and administrators must be developed. Concept maps are directed graph representations of conceptual relationships, and may be used to represent the content of a curriculum database. In this paper, we describe a Web application that uses Java-based concept maps was the user interface to a curriculum database.

World Wide Web-based access to heterogenous information resources for cytokine research and education.

Cytokines are currently the focus of intense research efforts, and information related to cytokine biology is expanding at an extremely rapid rate. The primary literature has become insufficient as a source for timely information, and efficient searches of the older literature are complicated by the nature of cytokine nomenclature. A few computer-accessible cytokine-related data sources have recently been developed, but these are typically limited in scope and not integrated with other information resources. We have developed Cytokine Explorer, a World Wide Web (WWW)-based resource to provide access to and guidance in use of information for research and education in cytokine biology. The resource consists of: (1) a searchable thesaurus of cytokine nomenclature; (2) a cytokine reagent database containing information from academic institutions and commercial sources; (3) a set of links to cytokine-related WWW sites; and (4) a database of functional information retrieval maps for guided retrieval of cytokine information from combinations of heterogeneous sources.

Virtual shelves in a digital library: a framework for access to networked information sources.

OBJECTIVE: Develop a framework for collections-based access to networked information sources that addresses the problem of location-dependent access to information sources. DESIGN: This framework uses a metaphor of a virtual shelf. A virtual shelf is a general-purpose server that is dedicated to a particular information subject class. The identifier of one of these servers identifies its subject class. Location-independent call numbers are assigned to information sources. Call numbers are based on standard vocabulary codes. The call numbers are first mapped to the location-independent identifiers of virtual shelves. When access to an information resource is required, a location directory provides a second mapping of these location-independent server identifiers to actual network locations. RESULTS: The framework has been implemented in two different systems. One system is based on the Open System Foundation/Distributed Computing Environment and the other is based on the World Wide Web. CONCLUSIONS: This framework applies in new ways traditional methods of library classification and cataloging. It is compatible with two traditional styles of selecting information searching and browsing. Traditional methods may be combined with new paradigms of information searching that will be able to take advantage of the special properties of digital information. Cooperation between the library-informational science community and the informatics community can provide a means for a continuing application of the knowledge and techniques of library science to the new problems of networked information sources.

Cost-effective clinical uses of wide-area networks: electronic mail as telemedicine.

Electronic mail (E-mail) is widely used as a means of communication in the medical community. E-mail is clearly inexpensive when compared to two-way, fully interactive, real-time, video telemedicine. By content analysis of 200 consecutive messages, we show E-mail to be a low-cost use of computer networks, supporting a wide range of physician decision-making.

Information retrieval maps for biomedical research.

Integration of computer-based information sources should be useful for biomedical research. An information retrieval map is a graphical representation of a set of expert rules for integrating computer-based information sources based on the data connections among those information sources. The use and implementation of an information retrieval map for sequence alignment is described.

Classifying and identifying servers for biomedical information retrieval.

Useful retrieval of biomedical information from network information sources requires methods for organized access to those information sources. This access must be organized in terms of the information content of information sources and in terms of the discovery of the network location of those information sources. We have developed an approach to providing organized access to information sources based on a scheme of hierarchical classifiers and identifiers of the servers providing access to those information sources. This approach uses MeSH tree numbers as both classifiers and identifiers of servers. MeSH tree numbers are used to indicate the information content of servers, and also as OSF/DCE server identifiers. This allows the identity and location of a server providing access to a given information source to be determined from the information classification of that information source.

OA-rehab: designing a personalized exercise program for people with osteoarthritis.

We describe the design of a multi-media performance support system (PSS) based on the documented benefits of a personalized exercise program for people with arthritis, on the known value of self-efficacy and stages of change, and on principles of learning theory. The poster will show examples of incorporating motivational and cognitive principles into a PSS.