The PRDBO organizes patient information into groups. Figure 9 shows the information groups of the PRDBO and how they relate to each other. The Identification Information Group contains information like name, address, and patient ID. The Demographic Information Group contains information like birth date and sex. The Encounter Information Group contains information like encounter date, physician seen, symptoms, and diagnosis. The Encounter Notes Group contains physician notes on the encounter. The Diagnostic Data Group contains the results of diagnostic procedures (e.g., X-rays) associated with the encounter. The Data Annotations Group contains annotations to the diagnostic data, such as, notations on an X-ray highlighting abnormalities. The Diagnostic Data and Data Annotations Groups usually contain multimedia information such as images and sound.
Pieces of information within a group have a one-to-one relationship to each other. For example, within the Demographic Information Group, each patient has only one birth date and is of only one sex.
The information groups can relate to each other in either a one-to-one relationship or a one-to-N relationship. For example, the Identification Information Group and the Demographic Information Group have a one-to-one relationship. Each patient has only one group of identification and demographic information. However, for each patient, there may be several visits to a physician. Consequently, there may be several Encounter Information Groups associated with each patient.
The information groups may be thought of as elements of sets. The PRDBO is a set whose elements are information groups or sets of information groups for each patient. Each element of the PRDBO set has as elements: the Identification Information Group, the Demographic Information Group, and a set of information about each encounter.
Figure 10 shows the architecture of the distributed application. Two clients of the PRDBO are being developed. One illustrates access from within the organization that created the information. This client is being developed using Object Linking and Embedding (OLE) on the PC. The other illustrates access from outside of the organization that created the data. This client is being developed for use with World Wide Web browsers. The arrows indicate the direction of patient information flow. The OLE application is capable of both reading and writing information to the data repositories within the organization which created the information. WWW browsers are capable only of reading information and they provide access to information created within an organization to those external to that organization.
The data repositories contain data suitable for traditional relational databases and multimedia data. The PRDBO Implementation and the OLE client are capable of performing SQL queries on relational databases. Where these databases are remote, the Remote Database Access (RDA) protocol with the SQL specialization (RDA/SQL) is used.
Multimedia data is best transmitted by means of the sockets interface of the Protocol Independent Interfaces IEEE Standard (PII/sockets). The PII/sockets interface is derived from the Berkeley sockets interface and is useful for transmitting large amounts of data.
The following sections will cover each of the two demonstration
projects in more detail.
