HL7 (Health Level Seven) RIM (Reference Information Model) version 1.0 was released in January 2001. HL7 version 3 messages to be developed in future are to be based on this HL7 RIM. HL7 version 3 messages are semantically regarded as a superset of HL7 version 2.x messages. Therefore, HL7 version 2.x messages, the most familiar healthcare information exchange protocol at present are naturally mapped to HL7 RIM. This leads to a guarantee of the interchangeability between HL7 version 2.x messages and version 3.x messages.
In addition to HL7 messages, J-MIX (The Japanese Set of Identifiers for Medical Record Exchange), MERIT-9 (MEdical Records, Images, Texts – Information eXchange), and MML (Medical Markup Language) are communication protocols that are often used for healthcare information exchanges in Japan. However, interchangeability among them is not yet trivial. Such a situation is a serious obstacle to implementing a large-scale healthcare information network in Japan. One of the solutions to build interchangeability is the development of a translator from one protocol to another. This solution requires at least 5C2 (=10) bi-directional protocol translators, development and maintenance of which would be labour intensive.
Our approach is to make HL7 RIM serve as an intermediate language for translation between the other protocols. Advantages of our approach are less cost for development (only 3 bi-directional protocol translators) and clear division of responsibility for development (each protocol developer being responsible for developing translating rules from/to HL7 RIM).
In this paper, we illustrate our strategy to map J-MIX, MERIT-9, and MML to HL7 RIM and demonstrate its effectiveness by examples of translating typical parts of the three protocols to HL7 RIM. This paper is the initial step in the process of unification of the existing healthcare information exchange protocols.
抄録全体を表示