Premium
Evaluation of the Roche cobas s 201 system and cobas TaqScreen multiplex test for blood screening: a European multicenter study
Author(s) -
Jarvis Lisa,
Becker Janine,
Tender Alzira,
Cleland Alexander,
Queiros Lucinda,
Aquiar Ana,
Azevedo Joana,
Aprili Giuseppe,
Bressan Fausto,
Torres Pilar,
Nieto Serafin,
Ursitti Antonella,
Montoro Jose,
Vila Enrique,
Ramada Concha,
Saldanha John
Publication year - 2008
Publication title -
transfusion
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.045
H-Index - 132
eISSN - 1537-2995
pISSN - 0041-1132
DOI - 10.1111/j.1537-2995.2008.01779.x
Subject(s) - multiplex , multicenter study , medicine , roche diagnostics , blood test , test (biology) , bioinformatics , biology , randomized controlled trial , paleontology
BACKGROUND: The Roche cobas TaqScreen test, an automated, multiplex nucleic acid test for blood screening for hepatitis B virus (HBV) DNA, hepatitis C virus (HCV) RNA, human immunodeficiency virus type 1 (HIV‐1) groups M and O, and HIV‐2 RNA, on the cobas s 201 platform, was evaluated by six European blood screening laboratories. STUDY DESIGN AND METHODS: The 95 percent limit of detection (LOD) of the cobas TaqScreen test for HBV, HCV, and HIV‐1, using dilutions of the WHO International Standards, were evaluated. The clinical performance was determined by testing between 2000 to 6000 routine donor samples. Some laboratories evaluated the robustness, cross‐contamination, and workflow. RESULTS: The mean 95 percent LOD (95% lower and upper confidence intervals) for HBV, HCV, and HIV‐1 across all the laboratories were 3.8 (range, 3.0‐5.2), 10.8 (range, 8.4‐14.4), and 56.7 (range, 43.0‐79.2) IU/mL, respectively. A total of 23,716 donors were tested in pools of 6. Fourteen initially reactive pools were detected, of which 6 contained a reactive donation, giving a positive predictive value of the pool results of 43 percent. One of the reactive donations was a HBV yield case (hepatitis B surface antigen–negative/anti‐HBc–positive). Evaluation of the workflow for the system showed that an optimized batch loading in which a pipettor (Hamilton Microlab Star IVD) was utilized to half capacity was better than a full batch loading. CONCLUSION: The 95 percent LOD for the three viruses were comparable to those obtained by Roche. The test and platform were shown to be sensitive, specific, flexible, and robust.