BACKGROUND: The operational and analytical performance of two automated triplex hepatitis B virus (HBV), hepatitis C virus (HCV), and human immunodeficiency virus (HIV) nucleic acid test (NAT) systems were compared in four screening laboratories of the French Blood Service.

STUDY DESIGN AND METHODS: Two laboratories evaluated the Procleix Tigris system (Chiron/Gen-Probe) in individual donation (ID) format and two sites used the cobas s 201 system (Roche Molecular Systems) on minipools (MPs) of six donations. The analytical sensitivity, the specificity, and operational performance were compared.

RESULTS: The ID to MP-NAT relative sensitivity factors in standard dilution panels of different genotypes varied between 8.7 and 21.9 for HCV RNA, 6.7 and 14.8 for HIV RNA, and 0.71 and 11.6 for HBV DNA. Tigris was 800-fold more sensitive than cobas s 201 (1:6) for a HIV group O sample, but did not detect the HIV-2 sample picked up by cobas s 201 with equal sensitivity as the HIV-1 group M samples. The specificity of both NAT systems after initial screening of 10,520 donations with Tigris and 1444 test pools on s 201 was 99.9 percent for both systems, but reached 100 percent after the repeat and pool resolution test algorithms. A higher throughput of the pool test protocol on cobas s 201 became apparent when the daily workload was more than 400 donations.

CONCLUSIONS: Tigris ID-NAT format was significantly more sensitive than cobas s 201 MP-NAT in detecting HCV RNA and HIV RNA dilution panels, but despite the 1:6 dilution factor in s 201 the difference in sensitivity was not significant for some of the HBV genotype panels. Both NAT systems demonstrated acceptable operational performance, but for routine use further improvement in system reliability is desirable.