Objective To evaluate the Carryover between the chemistry and immunoassay in Beckman Coulter Laboratory Au-tomation System and decide to whether sharing samples for testing between chemistry and immunoassay systems or not. Methods According to a certain order,high concentration samples and low concentration samples of HCG with different sample volume (500 μl,2 000 μl)were tested on Beckman AU5421 automatic biochemical analyzer.The HCG of low concen-tration samples were then tested to evaluate the carryover between the chemistry and immunoassay and explored the correc-tive procedure to deal with the carryover by increasing special cleaning process of beckman AU5421 automatic biochemistry analyzer.Results Under different sample volume,the carryover in a single module and as a whole of the beckman AU5421 automatic biochemistry analyzer were 5.44,15.47,23.51 and 45.96 ppm respectively (t=14.553,P <0.001;t=5.527,P =0.005;t=3.985,P =0.016;t=20.457,P <0.001).By increasing special cleaning process the carryover of 0.22 ppm was detected in 500 μl sample volume of the beckman AU5421 automatic biochemistry analyzer as a whole.Conclusion The car-ryover between the chemistry and immunoassay in Beckman Coulter Laboratory Automation System could been sovled by in-creasing special cleaning process of beckman AU5421 automatic biochemistry analyzer.

Objective To evaluate effects of 7 common hemoglobin variants on HbA 1c measurements using 4 ion exchange high performance liquid chromatography methods .Methods Ninety five samples with hemoglobin variants were collected from January 2017 to February 2018 during HbA1c measurements in laboratary medicine of peking university shenzhen hospital .Samples with 7 common hemoglobin variants were measured using Sebia Capillary 2 Flex Piercing, Bio-Rad D-10, Arkray HA8180V, Tosoh G8, and MQ6000 Plus, respectively.Effects of 7 common hemoglobin variants on HbA 1c measurements by the 4 methods were analyzed using Capillary 2 Flex Piercing as a comparative method .All statistical analyses were carried out using SPSS software version 19.0 .Mean bias were calculated for samples with hemoglobin variants , box plot was established to display bias distribution .Results Hb New York showed no interference on the 4 HPLC mechods although Hb New York could not be detected .D-10 could detect 6 Hb variants, and showed clinically significant interference for Hb J-Bangkok, Hb G-Coushatta, and Hb G-Taipei.HA-8180V fast mode yielded no HbA1c values for Hb J-Bangkok, Hb G-Coushatta, and Hb G-Taipei.Hb E, Hb Q-Thailand, and Hb G-Honolulu produced significant negative biases for HA-8180V.G8 standard mode could detect 1 Hb variant, and showed significant negative biases for six Hb variants .MQ6000 Plus could separate six Hb variants , only Hb G-Coushatta and Hb G-Taipei produced significant negative biases for the system . Conclusions Some common hemoglobin variants can interfere with HbA 1c determination by the most popular methods in South China , which may lead to erroneous HbA 1c values.