An Intelligent Web-Based Query System for Spatiotemporal Visualization of Infectious Diseases Surveillance Data in Taiwan

Abstract

Background: Beginning of 2005 Fall, the Taiwan Centers for Disease Control has activated a 3-year project to build an infectious disease data warehouse which obtains data with 15 different infectious diseases-related information systems. In order to better utilize these data and provide open public access, we have developed an integrated web-based query system featured in instant spatiotemporal analysis. Methods: For security concerns, a 4-tiered architecture was adopted, with an independent server, which is assigned to store the pre-calculated secondary data from the data warehouse on a daily basis. Database tables for default selected query and reports were reproduced everyday before dawn. An user online analytical and querying interface was created using Microsoft ASP.NET. Results: Visualization of notifiable disease surveillance data was the first developed web-based query system. Incidence trends are drawn on demand, which allow users to drill down the levels of geopolitical hierarchy, as well as different intervals and periods of time. Simultaneously, a swift Adobe Flash-embedded image map, instead of a clumsy geographic information system (GIS) server, provides users with geographical distribution of disease incidence accompanied with different colored legends. Automated year-to-year and age-group comparisons are also generated. The difference between traditional and intelligent web-based query system, new system saves at least 30 minutes for every query criterion and even saves the time required for raw data acquisition. Decreased paper-based surveillance reports and reduced budget on information distribution are anticipated. Conclusion: With the merit of this timely spatiotemporal analytic system, disease trends and spatial distribution can be visualized easily; it also saves time for creating these tables, figures, and maps. Furthermore, the system cuts down the learning curve for new epidemic analysts and reduce the budget of training on statistical and GIS software.