by Chau Vo, (Ho Chi Minh City University of Technology, Vietnam National University), Bao Ho (John von Neumann Institute, Vietnam National University) and Hung Son Nguyen, University of Warsaw
Inspired by MIMIC-III , VNUMED is a unified intermediate database of electronic medical records that is being developed in Vietnam. Its purpose is to gather medical records from hospitals, which can be used to support medical research.
Recent legislation mandates that every hospital in Vietnam must support electronic medical records . This is also encouraged by today’s Industry 4.0. To achieve this, a digital transformation of the medical field is required. This means electronic medical records must be established, in addition to the existing information system in each hospital. This came into effect in all hospitals under the Ministry of Health on March 1st, 2019, and all hospitals in Vietnam are required to have electronic medical records by the end of 2030. As a result, a huge number of electronic medical records are being generated and will be available in every hospital very soon. Compiling them thus lays the foundations for medical case-based research both within medicine and related fields.
Although new legislation  requires that we apply standardised technologies across hospitals , existing hospital information systems in Vietnam are very diverse, owing largely to differences in long-term investments in information technology among hospitals. Consequently, the development of electronic records has been a priority for some hospitals but not others. As outlined below, this presents huge challenges when it comes to using existing electronic medical records with external processing tools with the aim of gathering data to be used in research.
Firstly, the content of an electronic medical record needs to be well defined so that all the necessary details are available for reference in current treatment procedures and future processes. Traditionally, in Vietnam, like many other countries, hospitals have relied on paper medical records. Transferring all information from paper to electronic medical records is extremely difficult because of problems associated with understanding hand-writing, time pressure, computer skills, etc. In addition, records must be integrative so that not only their textual content but also their images and time series from medical tests are included.
To achieve this task, as part of our initial phase we investigated the use of the database MIMIC-III to support VNUMED. MIMIC-III is a popular database which is well processed and widely used, and we are considering both its database schema and practical applications for VNUMED. Such a choice also makes VNUMED independent of any electronic medical record type in any existing hospital information system in Vietnam. Furthermore, practical applications can then be constructed on VNUMED, hopefully like those on MIMIC-III.
Secondly, transferring electronic medical records in different hospital information systems into VNUMED is a big challenge, stemming not only from diverse information technologies, but also policies and connections between various organisations in medical and non-medical fields. This is a complex problem that relates to human as well as technical issues. Data integration always presents its own problems, but this situation is unique in that we are dealing with highly sensitive data relating to many patients and organisations.
Thirdly, such sensitive data must be well protected. In MIMIC-III, the rule-based de-identification method was used. For English data similar to those in MIMIC-III, other more effectively well-defined de-identification methods might offer potential. In our case, both English and Vietnamese data exist in VNUMED, thus, de-identification on VNUMED needs to be developed from scratch. Without an effective data protection scheme, VNUMED cannot be formed – and even if realised, VNUMED cannot be available for external research communities.
Last, but not least, once VNUMED gets started, post processing issues on VNUMED need to be taken into account for maintenance and general use. The first relates mainly to the internal development of VNUMED while the latter to external human users and application programs potential for VNUMED. Moreover, user-related policies need to be obtained for the latter.
Development of VNUMED is expected to be done step by step and every difficult aspect will be tackled as it arises. As soon as VNUMED is available, it will benefit researchers in a range of medical and non-medical fields. During the development of VNUMED, we are taking into consideration a range of possible uses, including electronic medical record visualisation, clinical text analysis, drug utilisation, disease diagnosis, etc. We anticipate that this work will contribute to the health and wellbeing of the Vietnamese people, and the international community.
In short, an intermediate database of electronic Vietnamese medical records, VNUMED, is being developed to provide valuable data for medical research. Many challenges lie ahead of VNUMED and we would appreciate any input and different perspectives that might help us achieve our goals.
This database is being built under a five-year research project funded by Vietnam National University at Ho Chi Minh City and the FIRST project of Ministry of Science and Technology, Vietnam.
 A. E.W. Johnson, et al.: “MIMIC-III, a freely accessible critical care database”, Sci. Data 3:160035, 2016. doi: 10.1038/sdata.2016.35
 Ministry of Health, Vietnam, Circular No. 46/2018/TT-BYT: “Regulations on Electronic Medical Records”, December 28, 2018.
 T. Benson and G. Grieve: “Principles of Health Interoperability: SNOMED CT, HL7 and FHIR” (3rd Edition), Springer, 2016. ISBN 978-3-319-30370-3 (eBook).
Hung Son Nguyen
University of Warsaw