Exchanging and integration of patient data across heterogeneous databases and institutional
boundaries offers many problems. We focused on two issues: (1) how to identify identical
patients between different systems and institutions while lacking universal patient
identifiers; and (2) how to link patient data across heterogeneous databases and institutional
boundaries. To solve these problems, we created a patient identification (ID) translation
model and a dynamic linking method in the Common Object Request Broker Architecture
(CORBA) environment. The algorithm for the patient ID translation is based on patient
attribute matching plus computer-based human checking; the method for dynamic linking
is temporal mapping. By implementing these methods into computer systems with help
of the distributed object computing technology, we built a prototype of a CORBA-based
object framework in which the patient ID translation and dynamic linking methods were
embedded. Our experiments with a Web-based user interface using the object framework
and dynamic linking through the object framework were successful. These methods are
important for exchanging and integrating patient data across heterogeneous databases
and institutional boundaries.
Keywords:
CORBA - Object Framework - Patient ID Translation - Dynamic Link - Exchange of Patient
Data
