The results of HbA_1c PT/ TPT organized by Shanghai Center for Clinical Laboratory (SCCL) from 2011 to 2014 were collected, and the 4-year datum change trend was analyzed about hospital and determination method distributions. The pass rates of each hospital and each method were analyzed by means of comparing the assigned values (target values) by the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) HbA_1c primary reference method and the group median values. Both College of American Pathologists (CAP) criteria and Shanghai local criteria of quality assessment for HbA_1c were also used. The performance (precision and accuracy) of each survey material and each group's pass rate were analyzed and compared with CAP global quality assessment results.

The number of participants increased from 245 in 2011 to 308 in 2014. The usage of domestic brand, MQ 2000PT, and imported brands, Bio-Rad, Tosoh and Arkray [high performance liquid chromatography (HPLC)], increased rapidly, and the rate was 38.1%-360.0%. A total of 3 point-of-care testing (POCT) devices were as follows: Axis-Shield (-78.9%), Boditech (no usage since 2013) and Quo-Test (usage since 2012, 27.8%). The usage rates of 2 low pressure liquid chromatography (LPLC) devices, Drew DS5 and MQ 2000, were -80.0% and -25.0%, respectively. The whole pass rate increased annually from 83.7% in 2011 (CAP 69.8%) to 95.8% in 2014 (CAP 90.3%). In 2014, the hospitals above Grade 2A had the pass rate 98.8%, while the CAP global quality assessment data were 88.8%-96.2% in the same year. The Grade 1 and Grade 2 hospitals had low pass rate than other grades. In terms of the total pass rate of each method group, it was higher by the use of group median values than by the use of assigned values of reference method, and also higher by Shanghai criteria than by CAP criteria. By CAP annual criteria, the pass rate of HPLC group still amounted to 81.0%-100.0%, except Tosoh in 2013, and those of 2 LPLC, 3 POCT and Immuno group (except in 2014) were 0%-75.0%. The total coefficient of variance (CV) of 5 survey materials/year was 3.0%- 4.0% for HbA_1c which improved obviously compared to 4.1%-9.9% in 2011 and be close to 3.4%-3.8% in CAP survey. The total bias was -0.16% HbA_1c-0.18% HbA_1c, which was lower than 0.3% HbA_1c criteria, but was higher than CAP's 0.02% HbA_1c--0.10% HbA_1c. Compared to the CAP 2 survey data with CV<3.5% (pass rate 85.0% and 90.0%), the pass rate in Shanghai was 83.3%, and the groups of POCT and LPLC methods cannot meet the minimum criteria of CV<5.0% and bias<0.5% HbA_1c.

By the 3-year quality management and TPT, the determination quality of HbA_1c is promoted, and the pass rates of hospitals above Grade 2A reach international level. Thus, the good performance products and qualified laboratories can assure HbA_1c as the diagnosis indicator of diabetes mellitus.