Analytical method transfer: New descriptive approach for acceptance criteria definition

Within the pharmaceutical industry, method transfers are now commonplace during the life cycle of an analytical method. Setting acceptance criteria for analytical transfers is, however, much more difficult than usually described. Criteria which are too wide may lead to the acceptance of a laboratory providing non-equivalent results, resulting in bad release/reject decisions for pharmaceutical products (a consumer risk). On the contrary, criteria which are too tight may lead to the rejection of an equivalent laboratory, resulting in time costs and delay in the transfer process (an industrial risk). The consumer risk has to be controlled first. But the risk does depend on the method capability (tolerance to method precision ratio). Analytical transfers were simulated for different scenarios (different method capabilities and transfer designs, 10,000 simulations per test). The results of the simulations showed that the method capability has a strong influence on the probability of success of its transfer. For the transfer design, the number of independent analytical runs to be performed on a same batch has much more influence than the number of replicates per run, especially when the inter-day variability of the method is high. A classic descriptive approach for analytical method transfer does not take into account the variability of the method, and therefore, no risks are controlled. Tools for designing analytical transfers and defining a new descriptive acceptance criterion, which take into account the intra- and inter-day variability of the method, are provided for a better risk evaluation by non-statisticians.