Jay Albert Brown*, Consultant, U.S. National Library of Medicine, Tacoma, WA, United States Track: Practice Presentation Topic: Online decision technology Presentation Type: Poster presentation Submission Type: Single Presentation Last modified: 2013-09-25

If you are the presenter of this abstract (or if you cite this abstract in a talk or on a poster), please show the QR code in your slide or poster (QR code contains this URL).

Introduction: In this age of information explosion, we need a way to index information for quick access to the facts so that we don't get lost in the details. Relational databases are commonly employed to organize company information, but not so widely used to map scientific knowledge. A relational database has many features that make it useful as a decision-support software tool. Most computer users are familiar with queries, a relatively simple and straightforward method of "zooming-in" on information. My objective is to show that an information-intensive knowledge domain (such as occupational toxicology or global infectious diseases) can be indexed into a relational database and serve as a decision-support tool to improve diagnosis and prevention.
Results: For the following two examples, the building process began within the conceptual framework of an intelligent database, that is, to first sketch the main features of the landscape and then to fill in the details.
"Haz-Map: A Relational Database of Hazardous Chemicals and Occupational Diseases" was started in 1991 and first published on the website of the National Library of Medicine in 2002. The goal of Haz-Map is to collect into one database the best information available regarding occupational exposures and diseases. Haz-Map was designed to improve access to information and to support the early recognition and prevention of work-related diseases. The seven main tables in the database are Chemicals, Industrial Processes, Home Activities, Occupational Diseases, Signs & Symptoms, Hazardous Job Tasks, and Occupations. All of the tables are linked so that queries can be performed to show all diseases that match a job AND a symptom or all chemicals that match an adverse effect AND an industrial process. There are 240 occupational diseases in the database. Each disease is linked to symptoms, hazardous job tasks, and causative chemicals. Since 2006 the number of chemical profiles in the database has increased from 1400 to greater than 10,000. In 2012, the website was "mobile enabled."
"IDdx: Infectious Disease Queries" was started in 2001 and is now available as a free iPhone, iPad, or Android App. The prototype application, containing 253 communicable diseases, was developed in Microsoft Access. Queries result in one or more diseases that match one or more search criteria. The disease search criteria include 99 signs & symptoms, 39 epidemiological factors, and 16 regions of the world. All information is bi-directional, i.e., the user can see all the symptoms associated with a disease or see all diseases associated with a symptom. The same structured vocabulary (indexing system) is used both to display information about a disease and to query the database. Each disease profile shows initial symptoms, incubation period, common findings, endemic areas, laboratory diagnostics, and unique epidemiological factors (entry, source, vector, and reservoir).
Conclusions: A continuously updated relational database, which includes or excludes information based on the needs of the users and depicts the details within the context of the whole map, is a powerful tool to support decisions regarding complex, medical knowledge domains.