【摘要】 目的 通过比较不同统计学方案对精液检验室间质评结果的影响，选择合适的统计学方案进行推广。方法 以精子浓度为例，选择2022年湖南省20家实验室精液室间质评数据，分别用传统统计方案、稳健统计方案以及结合稳健统计技术剔除“离群值”后的传统统计进行分析比对，比较其优劣。结果 4组精子浓度数据经传统统计无法剔除“离群值”,与Robust稳健统计方法或结合稳健统计技术剔除“离群值”后的传统统计室间质评结果有差异，传统统计与稳健统计合格实验室数分别为：19 vs 16、19 vs 16、19 vs 19、19 vs 19。结论 传统统计方法不适用于不符合正态分布、数据量较小、可能存在较多“离群值”的精液检验室间质评统计，推荐使用Robust稳健统计方法；或结合稳健统计技术剔除“离群值”后，可得到与稳健统计方案相似的结果。更多还原

【Abstract】 Objective To compare the effects of different statistical protocols on the results of external quality assessment(EQA) of semen, and select appropriate statistical protocols for the promotion of EQA of semen. Methods Taking sperm concentration as an example, the semen EQA data of 20 laboratories in Hunan Province in 2022 were selected, and the advantages and disadvantages of the traditional statistical scheme(TSS), robust statistical scheme(RSS) and traditional statistical scheme after eliminating the "outliers"(TSEOS) combined with robust statistical technology were analyzed and compared. Results The "outliers" could not be excluded from the sperm concentration data of the four groups in the TSS, which led to the difference between TSS and RSS or TSEOS. The number of qualified laboratories for TSS and RSS were 19 vs 16, 19 vs 16, 19 vs 19, and 19 vs 19, respectively. Conclusion The results of RSS are similar to those of TSEOS. Compared with TSS, RSS do not need to remove outlier data steps, and are more suitable for semen EQA data analysis with small data volume.更多还原