Quality by Design: Comparison of Design Space construction methods in the case of Design of Experiments

摘要

Quality by Design is a recent concept from quality control which has led to new requirements demanded by the authorities, particularly in the pharmaceutical industry. Among these, the guideline ICHQ8 explains that quality cannot be tested into products but should be built in by design. Designs of Experiments have been used to establish the relationship between input and output parameters of a process or formula while minimizing the risk in the final decision. More precisely, modelling studies are carried out to better understand the process and to predict the outputs in the whole domain of interest. When there are many input and output parameters, the experimenter looks for a compromise zone in which all the output parameters comply with the specifications. In this area, we can define a region where all the inputs can vary without altering the quality of the product with a fixed probability that the objectives will be reached. This zone is called Design Space and the determination of their boundaries is recommended by the Food and Drug Administration. In this publication, we propose different methods such as Monte-Carlo method, Bootstrap technique, numerical approach and reliability method to determine the Design Space. A comparison study is achieved using a case study.