Background: In external quality assessments (EQAs), it is important and necessary for participants to return authentic test results. However, to obtain better evaluation results, some laboratories have lacked confidence that their own and have preferred to rely upon the test results in surrounding laboratories. These violations covered up errors during testing and affected the authenticity of the returned data. The occurrence of violations in EQA launched by different EQA organizers is still unknown because these violations were carried out in private. In this paper, we adjusted the hepatitis B virus (HBV) qualitative EQA schemes to uncover some of the violations by using specially designed sample combinations in four EQA surveys to show the authenticity and accuracy of HBV testing in some prefectures of East China.
Methods: Four HBV qualitative EQA surveys were selected from 2016 to 2018, which were named 201602, 201701, 201702, and 201801, and a total of 474 laboratories were included. In the first two EQA surveys, the same EQA sample combinations were distributed. In the last two EQA surveys, three kinds of different sample combinations were designed and distributed with specific plans. The result of adding different sample combinations was that each laboratory only had a 33.3% chance of receiving the same sample combinations as those received by the surrounding laboratories. In addition, the laboratories were not aware that those samples with the same serial numbers might have different concentrations and target values. The sample concentrations, mode of delivery, requirements, results reports, and evaluation criteria were all the same as those of the first two EQA surveys.
Results: The number of laboratories with failed tests increased from fourfold to sevenfold between the first two and the last two EQA surveys. The number of failed tests in 201702 was eight times that in 201701 during the same year. Six (6 of 469, 1.28%) and three (3 of 472, 0.64%) poor performer laboratories (PPLabs) appeared in 201602 and 201701, respectively, while the number of those laboratories increased to 29 (29 of 474, 6.12%) in 201702 and 201801. Failed tests from PPLabs accounted for 70.5% (201702) and 68.7% (201801) of the total failed tests. The increase in the number of PPLabs was not universal but was concentrated in some prefectures.
Conclusions: The current EQA results for HBV qualitative testing were not as good as originally anticipated. Violations during the EQA surveys caused by exchanging test results with other laboratories and by modifying results before returning still occurred in some prefectures of East China. The laboratories that were in violation can be partly exposed as PPLabs by providing different kinds of sample combinations. Through such an EQA adjustment, the laboratory would have to re-establish confidence in its own testing results from HBV EQA samples. Then, evaluations of HBV qualitative testing samples in EQA would be more authentic and accurate in East China.
Keywords: authenticity; external quality assessment; qualitative testing; results exchange; target value; violations.
© 2019 Wiley Periodicals, Inc.