| Large repositories of health data are invaluable resources to researchers and planners in health care. These large repositories are usually created and maintained as data warehouses with data from many different sources being centralized on a large server. However, access to these repositories is often limited by the fact that specialized programs must be written only by programmers on the staff of the data repository who are familiar with the data themselves. This consultative process is both time-consuming and expensive.;An alternative access strategy is to make the data available on the Web, and have the end-user access the system through a Web-browser. The system guides the researcher through a set of questions so that the program is generated automatically by the system itself, thus bypassing the programmer bottleneck.;A prototype of such a Web-based access system has been developed for the Population Health Research Unit (PHRU).;The prototype system is based on three-tier architecture. The third tier is the PHRU data warehouse, and can still be used in the regular fashion. The first tier is the end user's Web browser. Through the browser, the user accesses only the middle tier, never the third tier directly. Applications developed in the middle tier help the researchers define the constraints of the application query, and then access the PHRU data on the third tier. On the other hand, they provide services to the PHRU staff to manage the system. (Abstract shortened by UMI.). |
