Assessment of Software Maintainability of openEHR Based Health Information Systems – A Case Study In Endoscopy

Maintaining health information systems over time requires significant effort and time. This is especially marked in clinical information systems where most, if not all, functional software requirements are dependent on healthcare concepts and processes which are prone to high rate of change. Software engineering literature indicates that maintenance tasks alone may constitute 70-80% of the total development cost. It has been suggested that openEHR based systems will effectively tackle this by separating domain knowledge from software code. The objective of this paper is to assess the maintainability of an openEHR based clinical application with comparison to another application based on the same functional requirements but implemented using traditional development methods. An endoscopy reporting application (GastrOS) driven by openEHR archetypes has been implemented using .Net/C#. It has the same functionality and appearance as an existing application which has been developed using Object/Procedural approach with relational data modelling. Afterwards a number of change requests have been implemented in both systems while assessing maintainability using metrics defined in the ISO/IEC 9216 and 25000 software quality standards. This paper presents the implementation methodology and preliminary results of the larger evaluation study using a more comprehensive set of change requests. These results indicate that, on average, the openEHR based application took approximately nine times less time to implement the change requests and were seven times less complex compared to the other application. While essentially a quantitative study it equally presents qualitative findings about opportunities and limitations of taking a model driven approach in development.

[1]  F M Zwiebel,et al.  Minimal standard terminology for digestive endoscopy: results of prospective testing and validation in the GASTER project. , 2000, Endoscopy.

[2]  Dipak Kalra,et al.  The openEHR Foundation. , 2005, Studies in health technology and informatics.

[3]  T Beale,et al.  Archetypes: Constraint-based Domain Models for Future-proof Information Systems , 2000, OOPSLA 2000.

[4]  Barbara Paech,et al.  Rational quality requirements for medical software , 2008, 2008 ACM/IEEE 30th International Conference on Software Engineering.

[5]  Ian Sommerville,et al.  Software engineering (6th ed.) , 2001 .

[6]  A. Rector Clinical Terminology: Why Is it so Hard? , 1999, Methods of Information in Medicine.

[7]  Koray Atala,et al.  Archetype based domain modeling for health information systems ; Sağlık enformasyon sistemleri için arketip tabanlı alan modellemesi , 2007 .

[8]  Minimal standard terminology in digestive endoscopy. , 2000, Endoscopy.

[9]  Magne Jørgensen,et al.  Assessing the Changeability of two Object-Oriented Design Alternatives--a Controlled Experiment , 2001, Empirical Software Engineering.

[10]  Lan Cao,et al.  Are Domain-Specific Models Easier to Maintain Than UML Models? , 2009, IEEE Software.

[11]  T Beale,et al.  The Health Record - why is it so hard? , 2005, Yearbook of medical informatics.

[12]  Grady Booch,et al.  Object-oriented analysis and design with applications, third edition , 2007, SOEN.

[13]  Robin C. Meili,et al.  Extrapolating evidence of health information technology savings and costs , 2005 .

[14]  Grady Booch,et al.  Object-Oriented Analysis and Design with Applications , 1990 .