Databases

A database is an integrated collection of information that is related to particular topic or purpose. Each database contains data elements such as patient name, medical record number, account numbers, addresses, phone numbers, and/or billing information. Databases allow you to organize, update, sort, and report on the information stored within them. Note that we identify a subtle difference between information stored in databases and spreadsheets: in speadsheets, the main purpose is to perform calculations with the data. Depending on the organization of the database and the specified software product that you are using, one or more files stored in the mass storage memory can make up a database. It is important to know how the database software of your choice stores the data on your computer since this has ramifications if you wish to copy the database to another computer.

We cannot escape jargon when we talk about databases. First of all, we briefly need to talk about how databases are organized and represented. A hierarchical database structure is how many companies and departments are organized.

Another structure is called a network (not to be confused with a computer network). In this representation, interconnections between the data are more easily represented, which allows for dependencies to be changed depending on the specific application or on the desired point of view.

Currently the most popular structure is the relational structure. In this approach, tables are used to represent the data. Each table is given a name and is called a relation. There can be many relations in one database. Commercial databases such as Microsoft's Access and Borland's Paradox are examples of relational databases. Because of the popularity of relational databases, we will illustrate the most important terms for this structure (note that various software products may use slightly different names):

· Field - a field contains one particular piece of information. In the table, a vertical column is a field: the patient's medical record number and name are two examples of fields. Fields can contain a number, an alphanumeric string, a date, a picture, or an entire 'blob' of text.

· Record - a record is a collection of fields that contain information about an entity in the table. In the table each horizontal row is a record: the patient's medical record number plus the patient's last name is one record.

Information in one table can be related to information in another table. This is an important concept because it enables the data to be organized into smaller and more manageable chunks that can be combined when needed. To illustrate this concept consider a database containing some demographic and some clinical information. One could decide to enter all these data into one table, but this involves much duplication of information. This makes it harder to make changes to the data and to keep the data consistent. Splitting the table into two separate, but related, tables greatly simplifies this by allowing you to work with one table at the time or with the two combined.

The software that manages databases is called a database management system (DBMS). A DBMS is defined as a system (both in terms of hardware and software) that lets you easily store, update, and retrieve information stored in the database. DBMS's have been available for many years, primarily only for use in large mainframe computers. Universities, large companies, and hospitals have used DBMS's successfully, provided they were able to scrape the funds together to acquire both the mainframe and the DBMS software. Nowadays, you can buy commercial, off-the-shelf DBMS software packages with essentially the same functionality as mainframe DBMS systems. These are available for virtually all types of computers: from desktop computers to computer networks to the large, water-cooled mainframes. For a PC the prices range from less than $100 for simple DBMS packages to about $500 for the more powerful desktop implementations, to a few thousand dollars for network-based DBMS. Tens of thousands of dollars will only get you started on the larger computer systems. The difference in price is mainly due to the fact that network and mainframe DBMS systems are multi-user systems, while the desktop DBMS is essentially a single-user system. Being a single-user system means that there is no need for implementing provisions for concurrent access and update to the same data elements; nor do these single-user packages have to deal with instances when hardware, network, or communication problems attempt to corrupt the integrity of the database. Typically, this responsibility falls onto the end-user (who may or may not be prepared to accept this responsibility.)

Contents

Prev. Page