Data Clearinghouse

Data warehouse forms the central utility at this level to provide a common architecture for integrating data extracted from operational subsystems and allows quick analysis of large, combined, complex data sets. Periodically, various rates, statistics, reports, forecasts, and health atlases are produced. Much of the information has to do with program evaluation, planning, forecasting, and managing. Conveniently, many of these statistical analyses, forecasting, and modeling may use the same or similar criteria, algorithms, and statistical models from time to time. They are repeatable and can be automated to a large extent in a data warehouse environment.

Transactional records of public health events are copied constantly into the data warehouses after their confidential components are removed. Their qualities are checked through statistical process control. Records are partitioned in the data warehouses based on time, demographic region, business function, and organizational unit to facilitate data manipulation and management. Other data sets such as demographic, social-economic, and environmental data are aggregated and geo-referenced in the same environment to allow joint spatial and statistical analysis. Through normalization and entity relationship modeling, data are integrated together and linked with predefined statistical algorithms, logic rules, and trend and forecasting models for information derivation. Data derived at this level are subject oriented, dynamic, and summarized with statistical certainty. They are snapshots and represent values over time.

The data warehouse constitutes the heart of the analytical processing of the hierarchical diagram. It serves the needs of planning, management, and evaluation of departmental programs in decision-making. A much smaller group, compared to those using the operational systems, accesses the data warehouse routinely to obtain necessary information.

Smaller in scope than data warehouse, a data mart is designed to meet the needs of a particular department. The PHSIS data mart is being developed since 2001. It will not only optimize current statistical analysis process, but also expand and enhance the capability of analysis by data linkage, data manipulations, etc. Data mart also lays an important ground to build Statistical Surveillance System.

Currently the population data, historic birth and death files are reorganized and stored in PHSIS data mart for internal use. In the near future the data sets will include cohort files, marriage, divorce, cancer data, PRAMS, etc, and its functionalities will grow too.