Disaster Planning and Management

University Of Pittsburgh At Pittsburgh, Pittsburgh PA

Investigators

Linked publications & trials

Abstract

This research will support the BioSense Initiative by developing a software platform for biosurveillance staff[unreadable] and researchers that allow them to measure the sensitivity, specificity, detection timeliness, and smallest[unreadable] detectable outbreak of a surveillance system. We refer to these properties as detectability characteristics.[unreadable] These measurements are critical for evaluating not only the performance of an overall system but also the[unreadable] performance of outbreak detection algorithms, studying the efficacy of surveillance data sources, and[unreadable] designing more effective surveillance systems.[unreadable] The research team has developed a prototype software application called HiFIDE v1.00 that allows users to[unreadable] assess the sizes and kinds of outbreaks that are detectable using surveillance data such as free-text triage[unreadable] chief complaints from emergency departments and sales of over-the-counter pharmaceutical. HiFIDE[unreadable] possesses a graphical user interface (GUI) that enables a user to examine the relationship between[unreadable] outbreak size, sensitivity, specificity, timeliness of detection, and the completeness of the surveillance data.[unreadable] The analysis is done using surveillance data for that jurisdiction.[unreadable] HiFIDE evaluates detectability for a jurisdiction by simulating surveillance data that it would observe during a[unreadable] real outbreak. HiFIDE forms this outbreak surveillance data by using real non-outbreak surveillance data for[unreadable] the jurisdiction and injecting the outbreak effect onto this data. HiFIDE v1.00 constructs the inject using data[unreadable] from a real outbreak in another jurisdiction.[unreadable] This proposal has the following specific aims:[unreadable] 1. To enhance HiFIDE by increasing the library of outbreaks and detection algorithms, improving analysis[unreadable] capability, and expanding features of the GUI.[unreadable] 2. To develop an Application Programming Interface (API) that enables HiFIDE to use algorithms,[unreadable] surveillance data, and injects from external surveillance systems to evaluate detectability for those systems.[unreadable] The communication between HiFIDE and an external system will be PHIN-compliant.[unreadable] 3. To evaluate the utility of HiFIDE for estimating outbreak features from a spike in surveillance data.