Long-term stability study of drug products and out-of-specification test results

A metrological approach for investigating out-of-specification (OOS) test results in long-term stability study of drug products was used. It is shown that OOS test results can indicate an actual change in a measured property of a product or be metrologically related with a certain confidence probability, i.e., be caused by the measurement problems, while the product still meets the quality requirements at the time of testing. As examples, results of testing sodium chloride injections in 500-mL plastic containers and of epinephrine (l-adrenaline) injections in 1-mL ampoules were discussed. Regression analysis of the data was performed, as well as warning and action lines for shelf life of the products calculated for relevant measurement uncertainties and confidence probabilities. Producer’s and consumer’s risks of the established shelf life values were also estimated.