The data of BD flow cytometry group from the National Center for Clinical Laboratory EQA in 2014 and Shanghai Center for Clinical Laboratory EQA for the first time in 2015 were analyzed statistically, and the standard deviation (s) and coefficient of variation (CV) were calculated. The internal quality control data from 2 laboratories in Shanghai were collected from September 2014 to April 2015, and the s and CV were calculated. The CD3⁺CD4⁺ percentage (CD3⁺CD4⁺%) data of Sample 201403 and Sample 201405 (target values: 42.7% and 42.1%) from the National Center for Clinical Laboratory, Sample 1521 and Sample 1525 (target values: 43.2% and 44.13%) from Shanghai Center for Clinical Laboratory and the submitted internal quality control data (monthly mean 44.39%- 46.80%) of 2 laboratories were analyzed comparatively.

The s and CV of 5 EQA samples from the National Center for Clinical Laboratory in 2014 were 1.10%-1.55% and 3.1%-5.5%, respectively, and the average CV was 3.36%. The s and CV of 5 EQA samples for Shanghai Center for Clinical Laboratory in 2015 were 0.67%-1.63% and 3.51%-8.64%, respectively, and the average CV was 4.83%. In the situation of data with approximate means, the group s (1.55% and 1.35%) and group CV (3.6% and 3.2%) from the National Center for Clinical Laboratory and the group s (1.63% and 1.55%) and group CV (3.78% and 3.51%) from Shanghai Center for Clinical Laboratory were less than those of internal quality control data from 2 laboratories (monthly s 1.06%-2.44% and 0.98%-2.03%; monthly CV 2.18%-5.28% and 2.14%-4.35%).

It is irrational that the s and CV of EQA are less than those of internal quality control since different impacts from the factors such as instrument brands and models, lysis techniques, brands and volumes of antibodies, gating strategies, experimenters and environment conditions and so on. One of the possible reasons may be that some laboratories exchanged, checked and sometime modified the EQA data before submitting the data. Other than improving the quality education for the laboratories, the solution can be the establishment of rational marking scheme.