The 131st Annual Meeting (November 15-19, 2003) of APHA

Abstract #70194

Zachary Pincus, David L Buckeridge, MD, MSc, Michael K Choy, Martin J O'Connor, MSc, and Mark A Musen, MD, PhD. Stanford Medical Informatics, Stanford University, MSOB X-215, 251 Campus Drive, Stanford, CA 94305-5479, (650) 725-3398, zpincus@stanford.edu

In an attempt to increase the timeliness of outbreak detection, many public health agencies are considering surveillance of non-clinical health data, such as school absenteeism and pharmacy sales. Using such data is difficult, however: Records from different sources may have divergent formats and semantic meanings, and cannot be accommodated by current standards for representing clinical data.

To facilitate the immediate use of non-clinical data in the absence of a comprehensive standard, we have created the BioSTORM Systems Abstraction and Interface Layer (BioSAIL), a flexible and extensible metadata model of such data and their sources. Part of our BioSTORM surveillance system, BioSAIL describes basic conceptual elements of non-clinical data and allows users to build detailed, customized descriptions of specific data sources and formats from these elements.

We have added a suite of mapping methods to BioSAIL that use its structured framework to transparently link raw data, from multiple sources and in arbitrary formats, with analytic methods. New data sources and analyses can be incorporated easily because the BioSAIL model and mappings provide a consistent interface between the two.

BioSAIL is in use integrating data from San Francisco 911 dispatches, a Veterans Affairs hospital, and computational disease models and simulations. BioSAIL collects these heterogeneous data, uses its metadata model to extract and link relevant information, and transfers the data to BioSTORM’s analytic methods in a uniform format with clear semantics.

BioSAIL allows rapid integration of varied data sources and epidemiological analyses in a modular manner, facilitating the deployment of syndromic surveillance systems.

Learning Objectives:

At the conclusion of the session, a participant in this session will be able to:
2. Assess the advantages and difficulties in incorporating non-clinical data into public health syndromic surveilance systems.
4. Analyze the requirements of and potential solutions for non-clinical data integration and utilization.
6. Evaluate the utility of knowledge-based (or "metadata-directed") systems in combining data from diverse sources.

Keywords: Bioterrorism, Surveillance

Related Web page: smi-web.stanford.edu/projects/biostorm/

Presenting author's disclosure statement:
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.