Data Collection /Sharing

Johns Hopkins University, Baltimore MD

Investigators

Abstract

This project will develop and evaluate a collection of public health technology tools to assist in[unreadable] remote data collection and dispersion. Mass prophylaxis clinic needs will be addressed to capture data[unreadable] electronically and monitor both individual and collections of clinics. Language translation features will be[unreadable] implemented to support real-time text-to-text or text-to-speech translation in the field, enabling clinic staff to[unreadable] communicate with speakers of other languages. Event reporting tools will allow health data from injured or[unreadable] sick individuals to be collected and uploaded to a surveillance system during large public events.[unreadable] Techniques using lightweight, low-investment, multi-use technologies will be readily adaptable to other[unreadable] population-based public health situations such as vaccine days and mass treatment and casualty[unreadable] situations.[unreadable] Using clustering, decision tree induction and other data-mining techniques we will develop[unreadable] automated, data-driven methods of defining syndrome groupings for syndromic surveillance systems. In[unreadable] addition we will compare the sensitivity and specificity of these data-generated syndromes with currently[unreadable] used expert-opinion based syndromes using the CDC's ICD-9-based syndrome definitions as the gold[unreadable] standard.[unreadable] We will extend available PHIN ontologies and vocabulary concepts to enable standardized[unreadable] interchange of aggregated and raw data, expert-user input text, and results of statistical tests within and[unreadable] between syndromic surveillance systems. The techniques developed will support the development of interjurisdictional[unreadable] detection algorithms in Project 2, and will provide standards for inter-jurisdictional transfer of[unreadable] information and aggregated data when sharing of raw data is legally or politically is not possible.[unreadable] The ability to collect, summarize and share data between diverse information systems is key to the[unreadable] creation of local and national public health informatics networks (PHIN). Although work will be centered on[unreadable] syndromic surveillance the techniques developed will be applicable to other areas of public health.