by Xiaohui Bi, Hongxia Ning, Tingting Wang, Dongdong Li, Yongming Liu, Tingfu Yang, Jiansheng Yu, Chuanmin Tao
The recent approval of 4th generation HIV tests has forced many laboratories to decide whether to shift from 3rd to these tests. There are limited published studies on the comparative evaluation of these two different assays. We compare the performance of fourth-generation electrochemiluminescence immunoassay (ChIA) and third-generation enzyme linked immunosorbent assay (EIA) for human immunodeficiency virus (HIV) screening and gauge whether the shift from EIA to ChIA could be better in a multiethnic region of China. Methodology/Principal Findings
We identified a large number of routine specimens (345,492) using two different assays from Jan 2008 to Aug 2011 in a teaching hospital with high sample throughput. Of the 344,596 specimens with interpretable HIV test results, 526(0.23%) of 228,761 using EIA and 303(0.26%) of 115,835 using ChIA were HIV-1 positive. The false-positive rate of EIA was lower than that of ChIA [0.03% vs. 0.08%, odds ratio 0.33 (95% confidence interval 0.24, 0.45)]. The positive predictive value (PPV) of EIA (89.6%) was significantly higher than that of ChIA (76.1%) (<0.001), reflecting the difference between the two assays. The clinical sensitivities of two assays in this study were 99.64% for EIA and 99.88% for ChIA. Conclusion
Caution is needed before shifting from 3rd to 4th generation HIV tests. Since none of these tests are perfect, different geographic and ethnic area probably require different considerations with regard to HIV testing methods, taking into account the local conditions.