Dear bravi,
there is no problem in preparing a matrix. I have made up one (which unfortunately I cannot send out, first because it contains proprietary information about our products and processes, and second because it is in spanish) based on a huge Excel file. Our approach is to select the tracer analyte for each product (basically the most soluble/most potent component of each formulation), and then the worst case among all the tracer analytes, in each piece of equipment. This to select for each piece of equipment the worst case contaminant, and the worksheet proposes this worst case based on the solubility and potency data, but the selection itself is performed manually in order to allow for decision taking on special cases (e.g. Vitamin E is the worst case, but a contamination with Vitamin E is far less risky than a contamination with Glimepiride even if Vitamin E is more insoluble and has a lower active dose than Glimepiride). Another worksheet sums the individual surfaces of each piece of equipment for each product to determine the surface of the equipment train, and a third worksheet (the core worksheet) performs a lookup function for each piece of equipment, each product and each tracer analyte. This worksheet calculates the factor [batch size/(maximum daily takings*equipment train surface)], selects the product which minimizes this factor (which is the worst case contaminated product) for each piece of equipment, and with the minimum daily dose of the worst case contaminating analyte calculates the maximum allowed contamination for each piece of equipment. It requires a lot of lookup function use, but can be done with little more than basic Excel knowledge. The disadvantage of this approach is that the contamination is calculated as a function of the whole equipment train and not for the shared equipment, but as this only lowers the maximum allowed contamination there is no formal issue other than needing somewhat more sensitive analytical methods. The number of products to consider depends on your specific case, but the only way to have only one worst case analyte is that every piece of equipment is dedicated except for the ones that are part of the equipment train where the worst case analyte is processed, which is highly improbable. The dispensing area needs not only to be considered in the equipment train but is one of the most critical parts, because it comes in contact with the pure API. Because of that, in the dispensing area we are using disposable scoops for the API’s, weigh the materia prima for only one product at a time and with the API’s as the last weighed components of the formula, and covering the balance plates with polyethylene sheets which after dispensing are discarded.
Best regards
Alfred
