GIDEON: a comprehensive Web-based resource for geographic medicine
© Berger; licensee BioMed Central Ltd. 2005
Received: 16 April 2005
Accepted: 22 April 2005
Published: 22 April 2005
GIDEON (Global Infectious Diseases and Epidemiology Network) is a web-based computer program designed for decision support and informatics in the field of Geographic Medicine. The first of four interactive modules generates a ranked differential diagnosis based on patient signs, symptoms, exposure history and country of disease acquisition. Additional options include syndromic disease surveillance capability and simulation of bioterrorism scenarios. The second module accesses detailed and current information regarding the status of 338 individual diseases in each of 220 countries. Over 50,000 disease images, maps and user-designed graphs may be downloaded for use in teaching and preparation of written materials. The third module is a comprehensive source on the use of 328 anti-infective drugs and vaccines, including a listing of over 9,500 international trade names. The fourth module can be used to characterize or identify any bacterium or yeast, based on laboratory phenotype. GIDEON is an up-to-date and comprehensive resource for Geographic Medicine.
As of 2005, the world is confronted by 338 generic infectious diseases, scattered in a complex fashion across over 220 countries and regions. Each new day confronts health care workers with unexpected outbreaks, epidemics and heretofore unknown pathogens. Over 2,000 named bacteria, viruses, fungi and parasites are known to cause human disease; and are confronted by 328 anti-infective agents and vaccines. Experts working in Health Geographics share an obvious and immediate need for comprehensive and timely data on the status of infection around the globe. A recent outline of GIDEON addressed uses for the Infectious Diseases clinician . This review will focus on the Global Health aspect of the program.
In 1990, we initiated a project to design computer systems to follow all diseases, outbreaks, pathogens and drugs. The initial DOS-based program was written in Paradox for floppy disks, later evolving through a compact disk-based program in Windows. A commercial web-based program was eventually released under the name, GIDEON (Global Infectious Diseases and Epidemiology ON-line, Gideon Informatics, Inc, Los Angeles, California) at http://www.GideonOnline.com. The current version is available on CD (updated every three months) or web subscription (updated every week).
The program consists of four modules: Diagnosis, Epidemiology, Therapy and Microbiology. Program modules of peripheral interest in Health Geographics (Therapy and Microbiology) will be discussed only briefly.
C = Chikungunya, P = probability or incidence, S = observed symptoms
P-(C/S) = probability of Chikungunya, given these symptoms
D2, D3, Dn = other diseases compabible with this clinical scenario
Two spreadsheets in the GIDEON database respectively follow the incidence of all symptoms for every disease, and the incidence of all diseases for every country. When a clinical case is "entered" into GIDEON, the program identifies all compatible diseases and ranks their relative likelihoods as determined by the above formula, ie: P-(C/S) vs. P-(D2/S) vs. P-(D3/S) ... vs. P-(Dn/S).
A blinded study of 500 cases conducted by this author found that the correct diagnosis was listed in the differential list in 94.7% of cases, and was ranked first in 75% . A second study of hospitalized patients in Boston found that the correct diagnosis was listed in only 69%, and was ranked first in 60% . It is likely that inclusion in the differential diagnosis list may be more important than disease ranking in such systems .
In theory, data entry by users can be monitored at the server level for purposes of surveillance. For example, if one or more users in China were to enter cases of fatal pneumonia, a "red-flag" at any monitoring agency (i.e., the World Health Organization) could indicate the possible appearance of SARS – long before submission of specimens or reporting of the case to local authorities. Similarly, the appearance of multiple cases of "dysentery" by users in a given community could indicate a possible outbreak of shigellosis.
The Therapy module follows the pharmacology and application of all drugs and vaccines used in Infectious Diseases. The current version contains 264 generic drugs and 64 vaccines. Various sub-modules present the mechanism of action; pharmacology, dosages, drug-drug interactions, contraindications, spectrum, and susceptibility testing standards. An international synonym lists contains over 9,500 trade names. As in other modules, users may add custom notes in their own language for each drug or vaccine: prices, resistance patterns, local trade names, etc.
The Microbiology option is similar to the Diagnosis module. Users may enter any combination of phenotypic tests, and obtain a ranked probability list of compatible bacteria. The current version incorporates more than 1,300 taxa. The Microbiology module is also designed to list the phenotype, prior names, ecology and disease association for any organism, or compare the phenotypes of any combination of organisms selected by the user.
Since the graphic and mapping functions of GIDEON treat individual countries as whole units, data presentations lack a certain degree of "granularity." Thus, the differential diagnosis of fever in Venezuela will include malaria, even if the patient is living outside of the endemic, southern region. This problem is corrected to a large extent by text in the associated country-specific notes and the general knowledge base of the treating physician. In theory, the manufacturer could follow the incidence of each disease for every state, district, province and oblast; but variability would still exist according to occupation, rural vs. urban setting, season, etc.
An additional problem relates to the availability and quality of valid epidemiological data. Disease reporting varies widely from country to country. For example, AIDS reporting statistics from sub-Saharan Africa are generally inadequate. Where necessary, the spreadsheets used by GIDEON record published true estimates rather than questionable reports. In other instances, Health Ministry data conflict with reports of the World Health Organisation, a fact which is recorded in relevant GIDEON country notes. Occasionally, major diseases are not reported at all. For example, several recent cases of cholera in Japan originated from Thailand; but Thailand has not officially reported a single case in many years. Where possible, the GIDEON data base relies on published best estimates, and at times 'educated guesses' when data are entirely lacking. Thus, there are few published data for disease incidence in Togo, and the program is forced to rely on publications for neighboring Ghana.
The reader is referred to the GIDEON website http://www.GideonOnline.com for an extensive listing of data sources, published reviews, technical background and pricing information.
- Edberg SC: Global infectious diseases and epidemiology network (GIDEON): A Web-based Program for Diagnosis and Informatics in Infectious Diseases. Clin Infect Dis. 2005, 40: 123-126. 10.1086/426549.PubMedView ArticleGoogle Scholar
- Berger S, Blackman U: Computer program for diagnosing and teaching geographic medicine. J Travel Med. 1995, 2: 199-203.PubMedView ArticleGoogle Scholar
- Ross JJ, Shapiro DS: Evaluation of the computer program GIDEON for the diagnosis of fever in patients admitted to a medical service. Clin Infect Dis. 1998, 26: 766-767.PubMedView ArticleGoogle Scholar
- Kassirer JP: A report card on computer-assisted diagnosis – the grade: C. N Engl J Med. 1994, 330: 1824-5. 10.1056/NEJM199406233302512.PubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.