A Product Line and Model Driven Approach for Interoperable EMR Messages Generation

A Health Information System (HIS) is a broad, integrated and standalone information system, designed to manage administrative, financial and medical records within a hospital and/or among hospitals. One of the fundamental parts of this system is called Electronic Medical Records (EMRs). EMRs are computerized legal medical records produced in hospitals. These records can conform to different medical messaging standards such as Health Level 7 (HL7) , HL7 CDA (Clinical Document Architecture) or openEHR (Electronic Health Records). In order to achieve the efficient transmission of all such messages across heterogeneous systems, in terms that each of them follows a different standard to record and store EMRs, such messages need to be interoperable and in XML format This problem may be solved by introducing a domain specific modeling language that defines the common and variable parts among various message standards (HL7, HL7 CDA and openEHR) using the concepts of Software Product Line Engineering (SPLE). Additionally, transformation rules that convert messages from one EMR standard to another is also one of the important features introduced. Our approach will help to build various messages that will conform to different EMR standards and may also help in achieving real-time requirements for the message transmission and reception. The contributions of this project are three-folds: (1) Users can specify EMR messages in the form of models, making them simple to design and potentially address evolution through model transformation techniques; (2) By applying SPLE concepts, a set of resultant messages sharing common and variable parts may be generated at once; and (3) iv The message generation can be extended to other domains as well, making it a more generalized application. v Preface The idea of this project came from the study of the messages generated in Health Information System (HIS). These messages, known as Electronic Medical Records (EMR), are presented in XML format and represent the critical and confidential information about a patient. Also, they can conform to different available medical message standards, such as HL7 and openEHR. The availability and speed of access of such messages can be very crucial in certain conditions. Also, a patient can have a number of EMRs, for example, records representing his demographic data or clinical observations. Trying to generate such messages one at a time can be a very cumbersome task. Additionally, messages that conform to different standards need to be made interoperable. A potential solution to this problem is to …