Search

Healthcare Information Systems (HIS) generally rely on large databases that store medical data. Apart from the specific issues related to the design of the data structures, these systems present a challenge to the developer regarding the architecture of the external level of the database, closely related to the building of the user interface. This paper discusses three specific issues related to the user interface of a medical database (The Temporal Property, The Allocation Process and The Open Attribute List Specialization) using a data-oriented approach for the external level. The approach is mostly based on data views, constructed on the server side and on the client application side. A small number of stored procedures however is necessary to implement the full functionality of the system. In our approach, most of the complexity of these processes is embedded in the queries that are run against the database. These queries can often be complex and time consuming. Two major issues arise from here: first, complexity can lead to inconsistent presentation of the same data (we are considering large information systems) and second, large queries can severely affect overall performance. These are the factors which the designers of such a system must control in order to reach an acceptable balance. By using the presented techniques, software developers can easily build user interfaces that present basic or medium-complex functionality, while having the benefits of standardization. For highly complex functionality, more advanced patterns and technologies are needed, which will be presented in a future paper. [ABSTRACT FROM AUTHOR]

In our past -work we have developed a data structuring technique based on trees and multitrees to model clinical research data. In this paper we try to assess the suitability of this model for keeping an electronic laboratory notebook by comparing its strengths and weaknesses with the products of some of the leading software providers on the market. The comparison focuses on five critical aspects: customizability, the use of tree-like structures, the use of patterns or templates, the techniques used to retrieve the data and functions that support collaboration. In the last two sections of the paper we discuss our findings and draw some conclusions that will serve as guidelines for our future work. [ABSTRACT FROM AUTHOR]