ABSTRACT
This chapter lays out a typical data request process in a health department. With the intention of building an agency-wide data warehouse, it discusses the requirements for the top-down and bottom-up approaches. The chapter describes some inefficient data linkage methods common in state public health agencies. Data requests are redundant because program-to-program data linkage tends to be piecemeal and ad hoc. The chapter presents the main ideas about data integration using the master patient index (MPI) and loosely joint programs and their datasets. In the process of exploring ways to integrate data, it provides examples of how to implement a bottom-up approach to build an MPI within the state public health agency. Based on examples and other practice-based data linkage projects, we loosely connected data marts or program data. The chapter concludes that the bottom-up approach is more appropriate for loosely connected and program-based public health departments.
