Quality by design in the pharmaceutical industry: Process modelling, monitoring and control using latent variable methods

The pharmaceutical industry has entered a new era. Attention is now being paid to real time process monitoring, real time process control, continuous improvement of processes, and quick product technology transfer. Terms like Quality by Design, Design Space, Control Strategy, Process Analytical technology, Process Signature reflect the current state. Multivariate Statistical Analysis has played an integral part in several industries, enabling process understanding, process monitoring, utilization of real time analysers and real time product release. It is therefore appropriate to see it as an integral part of the pharmaceutical industry effort to address issues like Design Space, Control Strategy, real time process signature monitoring, process understanding and correct technology transfer. In this work it is demonstrated that multivariate, data based statistical methods play a critical role in providing solutions to these issues. From determining the acceptability of raw material entering the plant to ensuring quality of the product that leaves the plant, the multivariate analysis philosophy should govern all the operations that take that raw material and convert it to a final product in a cost efficient way, while meeting safety and environmental constraints, from development to manufacturing to site transfer.