Du Ying, Sui Silian, Tan Dejiang, Zhang Heng, Chen Chen, He Qing, Chen Hua, Xiang Xinhua
Objective: To explore a more accurate and profound approach to distinguish the testing capabilities of laboratories by using three evaluation methods to evaluate the results of proficiency testing for the kinetic turbidimetric method of bacterial endotoxins. Methods: First, the results were collected from a round of proficiency testing program designed with repeated strategy of the kinetic turbidimetric method of bacterial endotoxins. Then, the assigned value and the proficiency testing standard deviation were calculated by using the robust algorithm A and robust algorithm S respectively under the logarithmic scale, and subsequently, three evaluation methods were utilized to assess the testing capabilities of the laboratories which were the quality acceptance criterion (ranging from 50% to 200%), the Z-scores, and the plots of ranges against averages. Results: After statistical processing of the test data under the logarithmic (log10) scale from forty laboratories, the assigned value X was 4.074 which obtained by using the robust algorithm A, and the proficiency testing standard deviation (σPT) was 0.042 by using the robust algorithm S. The lab capability,(1) When accessed by using the quality acceptance criterion method, the test results of all forty laboratories fell within the range of 50%-200% of the assigned value X, and thus were all categorized as satisfactory;(2) When accessed by using the Z-scores method: a. Thirty-six laboratories were rated as satisfactory, four as questionable, and none as unsatisfactory when using the mean of the intra-laboratory logarithmic results to calculate the Z-scores. b. Thirty-one laboratories were rated as satisfactory, six as questionable, and three as unsatisfactory when using each logarithmic result of the intra-laboratory to calculate the Z-scores; (3) When accessed by using the plots of ranges against averages method, twenty-five laboratories were rated as satisfactory, eight as questionable, and seven as unsatisfactory. Conclusion: Although the international community generally adopts the quality acceptance criterion method to make a final evaluation of whether the participating laboratories' capabilities (bacterial endotoxin proficiency testing) are satisfactory or not, it could be better distinguished the capability of the participating laboratories and could be more intuitively showed the bias from the assigned value and the degree of dispersion of each participating laboratory by the plots of ranges against averages method among these three evaluation methods.
