Abstract

The computer system used to process research data determines at least part of the reliability and quality of the data. For this reason, and also because guidelines like GCP, GLP, GMP and STERLAB are rather vague on computer systems, the Quality Assurance Unit and the Informatics Service Centre of the National Institute of Public Health and Environmental Protection took the initiative to create guidelines for computer systems. These guidelines support the laboratories of the institute to improve and maintain the quality of their computer systems. The project focussed on the quality of existing systems. The reasons are time restictions (a number of laboratories have to get certified in 1992), and the fact that there are more systems in existance than will be created this year. An information system consists of several parts. We call these parts "objects". Each object can be judged on several aspects. Because some aspects are relevant to more than one object, we built a matrix for the combination of objects and aspects. This matrix shows what we feel is essential for the quality of an information system. For a number of these combinations a guideline was created. This guideline consists of several diffrent pats: - a requirement - an explanation - how to evaluate compliance to the requirement. The guidelines are advisory. It is the laboratory that decides whether or not compliance is relevant for their system. A guideline can be ignored, but such a decision should be argumented and documented. With these guidelines we know what should be attained. The next part of the project will focus on how compliance to these guidelines can be achieved. Several software quality assurance systems exist. Furthermore, several laboratory quality systems exist. Unfortunately these systems are not easily applicable to our current sitiuation.