211718 Upgrading the ESSENCE Health Monitoring System for Increased Epidemiological Relevance and User Participation

Monday, November 9, 2009: 1:10 PM

Howard Burkom , National Security Technology Department, Johns Hopkins University, Laurel, MD
Yevgeniy Elbert , Johns Hopkins University Applied Physics Laboratory, Laurel, MD
Kenneth Cox , Armed Forces Health Surveillance Center, Silver Spring, MD
Joe Lombardo , Johns Hopkins University Applied Physics Laboratory, Laurel, MD
Current implementations of the Electronic Surveillance System for the Enhancement of Community-based Epidemics (ESSENCE) are limited in data sources employed and do not use data fields related to case severity. In some contexts, staff members monitoring many conditions and locations thus see nonspecific, sometimes excessive alerting, which reduces system usage. This effort exploits additional information from system data sources and more specific information from new sources to filter alerts adjustably by clinical relevance. This information is taken directly from disposition and evaluation codes, procedure codes and test orders and is also inferred from age distributions and temporal visit patterns. The goal is to combine statistical aberration detection and clinical knowledge according to available data characteristics to display anomalies with viewable criteria indicating that they should not be ignored. The validation approach is to evaluate criteria on large datasets with known outbreaks. The first study stage was restricted to outpatient visit records; subsequent stages include laboratory and radiology test results and filled prescription records. First-stage results were that alert flags are reduced by 50% at full service facilities and by much more at smaller ones. Severity criteria for remaining flags are readily inspected. In evaluation with datasets taken from five known outbreak datasets, detection timeliness and sensitivity were retained. The presentation will also discuss operational concepts for improved user training and acceptance.

Learning Objectives:
Analyze severity indicators from datasets of health records. Arrange statistical anomalies by clinical indicators. Evaluate surveillance system performance using historical data with effects of known outbreaks.

Presenting author's disclosure statement:

Qualified on the content I am responsible for because: I have worked as a developer and technical lead on ESSENCE biosurveillance systems for 10 years. For the past 2 years, my group has worked with the Environmental Protection Agency on development and implementation of a system module for the detection of waterborne disease outbreaks, and my role has been to work on the fusion of evidence, underlying algorithm development, and data analysis. I gave a related presentation at the 12th Biennial CDC/ATSDR Symposium on Statistical Methods, April 7-8, 2009.
Any relevant financial relationships? No

I agree to comply with the American Public Health Association Conflict of Interest and Commercial Support Guidelines, and to disclose to the participants any off-label or experimental uses of a commercial product or service discussed in my presentation.