It's How Medicine Should Be®

Translate

French German Italian Portuguese Russian

Guardian

What is GUARDIAN?

GUARDIAN (Geographic Utilization of Artificial Intelligence in Real-Time for Disease Identification and Alert Notification) is a real-time, automated, infectious agent detection and diagnosis system. GUARDIAN conducts real-time analysis of multiple pre-diagnostic parameters from records already being collected within an emergency department, such as nursing triage chief complaints, physician exam notes, radiology dictated reports and laboratory test results. GUARDIAN sends alerts to physicians' pagers notifying them of possible or confirmed cases of preprogrammed illnesses such as anthrax, smallpox, plague or even influenza the minute they are identified.  

About GUARDIAN

GUARDIAN was conceived and developed by Rush University Medical Center researchers who were introduced to disease surveillance during the 1995 Chicago heat disaster, from which they published work that evaluated the role of morbidity in predicting future mortality. GUARDIAN has received research funding from the Department of Defense Telemedicine and Advanced Technology Research Center dating back to 2006. (Contract Numbers:W81XWH-09-1-0662,W81XWH-06-1-0785)

Introduction: Real-time disease surveillance is critical for early detection of emerging infectious diseases, naturally occurring illnesses or the covert release of a biological threat agent. GUARDIAN is programmed to detect the spread of infectious agents by analyzing symptoms of patients as they are processed through the emergency department. More traditional trend analysis systems look at data collected and analyzed in a batch and sent to a lab — sometimes up to two weeks after the patient is seen. GUARDIAN analyzes the data in real time, meaning the test results are entered into the system and analyzed immediately. Using this technology, the system could potentially identify an outbreak of influenza or even an anthrax attack weeks in advance of traditional systems.

System: GUARDIAN is a real-time, scalable, extensible, automated, knowledge-based infectious agent detection and diagnosis system. GUARDIAN conducts real-time analysis of multiple prediagnostic parameters from records already being collected within an emergency department, such as nursing triage chief complaints, physician exam notes, radiology dictated reports and laboratory test results.

In addition to sending alerts to physicians' pagers about illnesses, the system is also able to geographically map where those cases have appeared. Together, these are powerful pieces of information that could help public health officials quickly contain an outbreak.

The GUARDIAN system currently consists of a pre-processor, an inference engine, an alert notification system, a human interaction system, a memory archiver and a set of relational databases managed by a database management system. A noteworthy feature of GUARDIAN is its ability to read and interpret free text utilizing a novel natural language processor.

Current project: GUARDIAN is undergoing extensive testing and validation particularly to measure its sensitivity, specificity and accuracy in monitoring patients for ILI per the CDC definition. DoD funding is being utilized to bring the advanced technology of GUARDIAN to other institutions for a more robust system of automated public health disease reporting.

Summary: GUARDIAN effectively balances the dual challenges of early detection of individual infectious agents and simultaneous detection of unusual patterns of disease occurrence in a target population. Using this system will assist clinicians in detecting potential BTAs as quickly and effectively as possible in order to better respond to and mitigate the effects of a large-scale outbreak.

GUARDIAN research team at Rush

  • Dino P. Rumoro, DO, FACEP
    Co-principal investigator
    Chairperson
    Associate professor
    Department of Emergency Medicine
  • Gordon Trenholme, MD
    James Lowenstine Professor of Medicine
    Director of the Division of Infectious Disease
    Rush University Medical Center
  • Shital Shah, PhD
    Assistant professor
    Health Systems Management
    Research scientist, Department of Emergency Medicine
  • Marilyn Hallock, MD
    Assistant professor
    Department of Emergency Medicine
  • Gillian Gibbs, MPH
    Project coordinator
    Department of Emergency Medicine
  • Kevin Thomas
    Information Services senior network analyst
    Department of Emergency Medicine

GUARDIAN partners

Pangaea Information Technologies, Ltd.

Publications and Presentations

Peer-reviewed articles related to the GUARDIAN project

  • Silva J, Shah S, Rumoro D, Bayram J, Hallock M, Gibbs G, Waddell M. Comparing the accuracy of syndrome surveillance systems in detecting influenza-like illness: GUARDIAN vs. RODS vs. electronic medical record reports. Artificial Intelligence in Medicine. 2013;59:169-174.
  • Rumoro D, Bayram J, Silva J, Shah S, Hallock M, Gibbs G, Waddell M. The impact of alternative diagnoses on the utility of influenza-like illness case definition to detect the 2009 H1N1 pandemic. American Journal of Disaster Medicine. 2012;7:105-110.

Published abstracts related to the GUARDIAN project

  • Rumoro D, Hallock M, Silva J, Shah S, Gibbs G, Trenholme G, Waddell M. Why Does Influenza-like Illness Surveillance Miss True Influenza Cases in the Emergency Department? Implications for Healthcare Providers. Annals of Emergency Medicine. 2013; 62(4):S75.
  • Silva J, Shah S, Rumoro D, Hallock M, Gibbs G, Waddell M. A Novel Syndrome Definition Validation Approach for Rarely Occurring Diseases. Online Journal of Public Health Informatics. 2013; 5(1).
  • Silva J, Rumoro D, Shah S, Gibbs G, Hallock M, Waddell M, Doseck S. Adaptation of GUARDIAN for Syndromic Surveillance During the NATO Summit. Online Journal of Public Health Informatics. 2013; 5(1).
  • Rumoro D, Silva J, Hallock M, Shah S, Gibbs G, Waddell M. Disease model fitness and threshold creation for surveillance of infectious diseases. Emerging Health Threats Journal. 2011;4:11125.
  • Rumoro D, Shah S, Silva J, Hallock M, Gibbs G, Waddell M. Case definition for real-time surveillance of influenza-like illness. Emerging Health Threats Journal. 2011;4:11123.
  • Silva J, Rumoro D, Hallock M, Shah S, Gibbs G, Waddell M. Disease profile development methodology for Syndromic surveillance of biological threat agents. Emerging Health Threats Journal. 2011;4:11129
  • Rumoro D, Silva J, Shah S, Bayram J, Hallock M, Gibbs G, Waddell M, Thomas K. Impact of Alternate Diagnoses on the Accuracy of Influenza-like Illness Case Definition Used for H1N1 Screening in the Emergency Department. Emerging Health Threats Journal. 2011;4:s110.
  • Rumoro D, Shah S, Gibbs G, Silva J, Bayram J, Hallock M, Waddell M. The impact of specific surveillance system methodology on influenza-like illness prevalence rates. Annals of Emergency Medicine. 2010;56:S6.
  • Waddell M, Meraz C, Silva J, Rumoro D. Application of Natural Language Parsers To Syndromic Surveillance. Advances in Disease Surveillance. 2008;5:71. http://isdsjournal.org//articles/3278.pdf
  • Silva J, Rumoro D, Waddell M, Doseck S. The Development of BTA-Specific Disease Profiles for Use in a Real-Time Disease Identification and Notification System. Advances in Disease Surveillance. 2007;4:16. http://isdsjournal.org//articles/1948.pdf
  • Waddell M, Doseck S, Silva J, Rumoro D. GUARDIAN: Geographic Utilization of Artificial intelligence in Real-Time for Disease Identification and Notification. Advances in Disease Surveillance. 2007;4:63. http://www.isdsjournal.org//articles/1975.pdf

Presentations on the GUARDIAN project

  • Silva J, Rumoro D, Bayram J, Shah S, Hallock M, Gibbs G, Waddell M. Comparison of Multiple Syndrome Surveillance Programs to Detect Influenza-like Illness, Poster Presentation at American Medical Informatics Association Annual Conference, November 2010.
  • Silva J, Hallock M, Rumoro D, Shah S, Gibbs G, Bayram J, Waddell M. Disease Surveillance: The Need for a Robust Natural Language Processor, Round Table Discussion at American Public Health Association Annual Conference, November 2010. http://apha.confex.com/apha/138am/webprogram/Paper221298.html
  • Rumoro D, Bayram J, Silva J, Shah S, Hallock M, Gibbs G, Waddell M. Applicability of Influenza-Like Illness Surveillance Guidelines for H1N1 Screening, Poster Presentation at American Public Health Association Annual Conference, November 2010.