RE: which statistical method could be used in process validation
Contribution from Dick Bonner written on 4/12/2012 at 13:02
Original contribution: which statistical method could be used in process validation written by hello on 4/6/2012 at 04:31)
I am afraid that it is not possible to give you a definitive answer. Each part of the process needs to be validated, for instance a mixing step should be tested for homogeneity so the number of samples taken will depend on the size and shape of the mixer, and the sample size tested will depend on what the next step of the process is. For instance if the mixing was to be followed by a tableting step the sample size to be tested should be no bigger than twice the tablet weight and the acceptance criteria would be the same as that used for weight uniformity testing of the tablets. I would recommend you produce a process flow chart of the process to be validated and then review what the critical process parameters are of each step, and from there determine what needs to be sampled and tested. The key to success is to identify what the acceptance criteria is before running the validation and keeping them as simple as possible.
Answering this Contribution: