• Military
• WMD
• Intelligence

• Security Menu                     

• Systems
• Industry
• Operations
• Hot Documents
• News
• Reports
• Policy
• Budget
• Congress
• Links

• Space

Early Aberration Reporting System (EARS)

The Early Aberration Reporting System (EARS) was pioneered as a method for monitoring bioterrorism during large-scale events. Its evolution to a standard surveillance tool began in New York City and the national capitol region following the terrorist attacks of September 11, 2001.

The primary purpose of EARS is to provide national, state, and local health departments with several alternative aberration detection methods. EARS helps assist local and state health officials to focus limited resources on appropriate activities during epidemiological investigations of important public health events. Finally, EARS allows end users to select validated aberration detection methods and modify sensitivity and specificity thresholds to values considered to be of public health importance by local and state health departments.

Various city, county, and state public health officials in the United States and abroad currently use EARS on syndromic data from emergency departments, 911 calls, physician office data, school and business absenteeism, and over-the-counter drug sales. EARS is convenient, easy to use, and available at no cost.

The EARS program presents its analysis in a complete HTML Website containing tables and graphs linked through a home page. Viewing EARS output requires only a Web browser.

EARS consists of a class of quality- control (QC) charts, including Shewhart chart (P-chart), moving average (MA), and variations of cumulative sum (CUSUM). Many syndromic surveillance systems use EARS for temporal aberration detection; some also use other QC charts such as exponentially weighted moving average (EWMA). A common characteristic in adopting these QC charts for syndromic data analysis is the use of a sample estimate for the baseline mean and standard deviation (SD). This approach circumvents the difficulties associated with the modeling of the baseline trend of the syndrome, a process complicated by the discreteness, serial correlation, seasonality, and daily fluctuation of the syndromic data. At present, understanding of these methods within the context of syndromic data is limited, and systematic evaluations of syndromic surveillance have not been conducted.

The QC charts in EARS use daily syndromic counts or incidences (daily counts of a specific syndrome divided by total ED volume for the day) between a past date and the current date to derive a monitoring statistic (mt). QC charts are traditionally evaluated with respect to the sensitivity and false alarm rate (one minus specificity) of single day detections. Because disease outbreaks probably are associated with temporal patterns, corresponding patterns of aberration signals should be considered. For example, with a disease outbreak that persists at a high level of incidence or count for a number of days after onset, consecutive signals might alert not only the onset but also the duration of the outbreak. Such a pattern of signals is called a detection event. More importantly, public health workers can use a detection event to estimate the duration or strength of an outbreak and respond accordingly. Therefore, detection events of composite detection signals (e.g., the first or the first pair of consecutive signals) are used to define sensitivity and false alarm rates, which are in turn converted to ROC curves and used as an overall performance measure.