Validating a method of analysis goes through different steps, which aims at testing the normality of measurements distribution, estimating the uncertainty of the components of a measurement （i.e., accuracy and correc...Validating a method of analysis goes through different steps, which aims at testing the normality of measurements distribution, estimating the uncertainty of the components of a measurement （i.e., accuracy and correctness）, and finally, define the control tests of non degradation of the method performances. This paper outlines the steps for validating a biological method of analysis. It involves the construction of an experimental design, a statistical model, and the preparation of an interne laboratory reference material （pilot vaccine）. The latter is used to study the impact of deviation and variation factors, in order to, optimize the analytical method, to evaluate the bias （random error）, and to calculate the uncertainty of measurement, and make the control charts. This method is applied in the titration of live viral vaccines of Gumboro disease on chicken＇s embryos fibroblasts. The experimental results show that potential influence factors related to the titration method had no significant influence on the obtained results. Taking into account these results, an operating mode has been elaborated. The finalized method proved to be faithful to standard deviation of repeatability and reproducibility of 0.21 and 0.22, respectively, with a confidence level of 95%. The calculated uncertainty of measurement is equal to 0.2, which represents the average error level of a titer. A homogeneous stock of interne laboratory reference vaccine （MRIL）, with an average titer of 5.9 log DIT 50, was produced and the control chart set in away to provide the laboratory with an important tool of control and monitoring of the viral titers evolution in time, as well as, the mastery of the validated titration method performances.展开更多
文摘Validating a method of analysis goes through different steps, which aims at testing the normality of measurements distribution, estimating the uncertainty of the components of a measurement （i.e., accuracy and correctness）, and finally, define the control tests of non degradation of the method performances. This paper outlines the steps for validating a biological method of analysis. It involves the construction of an experimental design, a statistical model, and the preparation of an interne laboratory reference material （pilot vaccine）. The latter is used to study the impact of deviation and variation factors, in order to, optimize the analytical method, to evaluate the bias （random error）, and to calculate the uncertainty of measurement, and make the control charts. This method is applied in the titration of live viral vaccines of Gumboro disease on chicken＇s embryos fibroblasts. The experimental results show that potential influence factors related to the titration method had no significant influence on the obtained results. Taking into account these results, an operating mode has been elaborated. The finalized method proved to be faithful to standard deviation of repeatability and reproducibility of 0.21 and 0.22, respectively, with a confidence level of 95%. The calculated uncertainty of measurement is equal to 0.2, which represents the average error level of a titer. A homogeneous stock of interne laboratory reference vaccine （MRIL）, with an average titer of 5.9 log DIT 50, was produced and the control chart set in away to provide the laboratory with an important tool of control and monitoring of the viral titers evolution in time, as well as, the mastery of the validated titration method performances.
