Database

Database Organization

The basic element of a database is a field, or variable. Each field in a database is specified as a fixed (maximum) number of characters, each equivalent to a byte of data. The data elements may be text, such as a patient name, or numeric, such as a birth date. Text data also can be coded, or converted to numeric values. For example, in a diagnostic database the value 250.0 represents a diagnosis of diabetes mellitus, using the International Classification of Diseases coding system. Specifications for the number of characters, the type of data, code definitions for all fields, data relationships, and other data structure characteristics are stored in a centralized repository called the data dictionary. The data dictionary is a vital part of the database management system.

A group of related fields is called a record. Typically a record contains data related to a person, location, or time-based event, such as an episode of care. Fields including patient demographics, diagnostic and procedure codes, and charge data might constitute that patient's financial record. The financial records of all patients would comprise the billing file, a collection of related records. A spreadsheet program such as Microsoft Excel can be used to construct a simplistic row-by-column “flat” file. In this structure, each cell on the worksheet is a field, and the columns show all observed values for a given variable. Rows are equivalent to records. Billing files, cost accounting files, insurance benefits files, and other related files could constitute a financial database, a pool of related and integrated data.

Data Models

Data stored in the database are organized in logical and physical configurations. The physical design, which determines where individual data elements are recorded on the storage medium, is the purview of computer technicians. The logical structure is conceptual, specifying data relationships used to group related variables. For example, all data elements needed to process a patient bill would be grouped. This design element determines the ultimate utility of the database. Thus involving end users of the database in the logical design process is critical.

The individual data elements and their relationships may be visually specified in a data model, a diagram similar to a flowchart, that shows the relationships and levels of data. A hierarchical model is a branching tree structure, where each level represents all possible elements of the level above it. For example, the personnel database might begin with all job classifications, one of which is nursing. The nursing classification might branch to nurse practitioners (NPs), registered nurses (RNs), and licensed practical nurses (LPNs). The RNs might branch to employment status (full-time regulars, part-time floaters, and temporary contract employees) and clinical assignments (medical, special care, postsurgical, psychiatric). This branching would continue to the individual employee level. Only one path exists from the starting point to the final element.

...