Abstract: Objective To discuss the comprehensive evaluation of the correlation between Enzyme-linked immunosorbent assay （ELISA）and nucleic acid test （NAT）in decreasing the risk of transfusion-associated infection was carried out via the detection model of ELISA in parallel with NAT,which in turn was combined with the results from confirmatory assays. Methods Positive samples from screening assays between May 2015 to July 2016 were analyzed in terms of the S/COvalues of two kinds of ELISA reagents as well as results of both NAT and Western blotting,the latter being confirmatory. Results The results of all 281 tested samples were as followed,in primary screening test：110 cases of Wantai reagent withS/CO≥1 and 156 cases of BIO-RAD reagent; 20 cases of Wantai reagent with 0.75<S/CO<1,39 cases from BIO-RAD reagent. HIV confiormatoryive WB test：27 were positive and 242 were negative,49 cases could not be determined by this method. NAT test：49 were positive and 232 were negative. Conclusion The reaction rate of two kinds of ELISA screening reagents was different,the difference was statistically significant. Results showing strong reactivity in ELISA had higher consistency with NAT whilst low ELISA reactivity samples delivered lower consistency with NAT and Western blotting. Besides,conducting ELISA for one less time leads to higher requirements on both reagents and operating personnel. At the same time,we should strengthen recruitment and promotion,strengthening publicity of high-risk sexual behavior people voluntarily gave up blood donation of medical purposes,established fixed blood donor of blood donor recruitment is more conducive to blood safety.

Key words: HIV, ELISA, Western blotting, NAT