• Japanese law requires doctors to "write down" the contents of medical care. This could force them to use and keep the paper chart (until last March).
• Physicians themselves perform data and order entry at the point-of-care. This is especially noteworthy because in Japan there are no physician’s assistants, the duties of nurses are limited, and most hospitals do not have clerks in patient care areas.
• Difficulty with input of Japanese Characters into the computer.
• In the past, the patient-doctor encounter was so short that patients complained of a "three hour wait for a three minute encounter." While this situation has improved, physicians have very little time for data/order entry.

Thus doctors tend to enter items (like prescriptions) that produce money, and tend to neglect others (like signs and symptoms) that are essential to clinical decision making.

Pressed to cut costs without lowering the quality of medical care, government authorities are looking to the CPR to support the use of diagnosis-related groups and a prospective payment system as well as the use of system audits of the appropriateness of medical care/services. To meet these goals, authorities gave approval to the CPR in April 1999.

Among so called CPR systems, many of Japan's current patient-centered order-entry systems lack certain desirable CPR characteristics or functions in its database design. The necessity for integration (or the reciprocal semantic links) of patient data/information which is accumulated among sub-systems is often ignored. They can, however, form the base for a stepwise transition to a CPR system. With some hospitals launching efforts to design, develop, and/or examine electronic problem oriented medical record systems, the transition is well underway.

Yet another problem to be resolved is the fact that most hospitals in Japan have single vendor systems and their HIS are based on mainframe and COBOL. They are not true open systems, at least at the level of application interfaces and data format. Without modules for inter-system communication, transactions still concentrate on the mainframe, and transaction control intermingles with database management. The transition to the CPR requires middleware to construct a true distribution system, distribute the load of transaction, and ensure the independence of clients and servers.

The transition also involves extensive work with terminology and nomenclature: this work is complicated by the nature of the Japanese language with its Kanjis (ideographical letters or so-called Chinese characters) and Kanas (phonographical letters of Japan original).

Transition from order-entry systems to the CPR will take major hospitals about six years. Near term, within the year, the Japan Association of Medical Informatics (JAMI) and the Japanese Association of Health Information Systems Industry (JAHIS) are working on

• Guidelines for security policies and regulations
• Standardization in the exchange of clinical data and information

• MERIT-IX (Medical Record, Image, Text - Information eXchange): XML, HL7, DICOM3 and/or MIME based exchange format & viewer.
• MML (Medical markup language): SGML based chart structure definition.
• JLAC: lab test code systems developed by Japan Society of Clinical Pathology

MERIT-IX has decided to adopt JLAC. Since JLAC9 (rev. in 1996) has a multi-axial structure, it is compatible with LOINC.

One of working groups of JAMI is also working with object modeling and object broker strategies. JAMI, MEDIS-DC(Medical Information System Development Center) and Ministry of Health and Welfare are working with the standardization of terms/nomenclature in Japanese language and of code systems. These will continue over the next 3 to 5 years, if there is adequate research funding. Other issues to be addressed include:

• Standardization of the data warehouse schema design
• Guidelines for system audit applications, i.e., confidentiality/security and appropriateness of clinical procedures