Design of Experiments

Design of Experiments (DoE) is a structured and efficient approach to experimentation which employs statistical techniques to investigate potentially significant factors and determine their cause-and-effect relationship on the outcome of an experiment.[1][2][3] Where a relationship between the experimental parameters (factors) and the results exists, the correlation can be detected and quantified. Hence, DoE can be applied to any process for which the inputs (factors or parameters) can be controlled, and the outputs (results) can be measured in a statistically valid manner.  DoE can be employed to design optimal and robust processes.  

Due to this, DoE has been widely used for many years and across many industries, sectors and disciplines including the pharmaceutical industry. In the pharma industry, it is most frequently used in the development and scale-up of new processes for the manufacture of intermediates for potential new drugs (once a decision has been made on the synthetic route).[4][5] Often, it is also used in the formulation processes required for the final active pharmaceutical ingredient (API).[6] After a new drug is successfully launched, it can be employed in the initial manufacturing processes to further optimise yield and quality and reduce manufacturing costs (although these studies are rarely published).

This module on Design of Experiments has been prepared by Sandrine Olazabal from GlaxoSmithKline.

Learning Objectives:

By the end of this module you should:

• Be familiar with the steps to follow when designing an experimental design;
• Understand how to identify main effects and interaction effects within a process;
• Understand how to DoE can be employed to discern settings to deliver acceptable and optimum performance;
• Be aware of how to control variation in a DoE execution;
• Be aware of the different types of experimental design approaches and their associated advantages and limitations.