First international proficiency study on West Nile virus molecular detection.

BACKGROUND West Nile virus (WNV) molecular detection is being conducted by a growing number of laboratories, but the degree of proficiency may vary between them. External quality control is needed. METHODS We have conducted an international quality assurance study on WNV molecular detection. Participating laboratories tested noninfectious samples inactivated by heat and gamma irradiation. Participants received 7 coded lyophilized samples containing WNV of genetic lineages 1a, 1b, and 2 at 2600 to 18 000 000 RNA copies/mL, 3 samples containing heterologous flaviviruses, and 2 negative samples. RESULTS Thirty laboratories participated. The average laboratory achieved 50% detection probability from 7762 copies/mL onward (probit analysis; 95% CI = 1174-24547 copies/mL). Lineages 1a and 1b were detected with equal efficiencies, but the lineage 2 strain (Ug37) was detected at significantly lower rates. Only 27% of participants were able to detect the 6 samples containing > or =1.8 x 10(4) copies/mL. Three laboratories generated false-positive results in negative samples. Six of 30 laboratories reported correct strain identification in 3 samples containing non-WNV flaviviruses. We observed a significant positive correlation between the capability of detecting non-WNV flaviviruses and detecting WNV lineage 2. CONCLUSIONS Most participants showed good performance in detecting lineage 1 WNV, the predominant virus in the Northern Hemisphere. The inability of some laboratories to detect even highly concentrated lineage 2 WNV downgraded the overall outcome. The lineage 2 material received through this study will provide laboratories with the necessary template for improving their assays. Such material is otherwise hard to obtain.