Melting curve analysis reveals false-positive norovirus detection in a molecular syndromic panel

Abstract

AbstractBackgroundMolecular syndromic panels can improve rapidity of results and ease clinical laboratory workflow, although caution has been raised for potential false-positive results. Upon implementation of a new panel for infectious diarrhea (BioFire® FilmArray® Gastrointestinal [GI] Panel, bioMérieux) in our clinical laboratory, a high number of stool samples with norovirus were detected.ObjectivesTo investigate positive percent agreement and the false-positive rate of norovirus detected by the multiplex Biofire GI panel compared to a singleplex commercial assay.Study designFrom October 2023 to January 2024, all prospective stool samples with a positive norovirus result by BioFire had melting curves reviewed manually using the BioFire FilmArray Torch System. Stool samples further underwent testing by a supplementary real-time RT-PCR assay (Xpert® Norovirus, Cepheid) for comparative analysis.ResultsOf the 50 stool samples with norovirus detected by BioFire, 18 (36%) tested negative by Xpert (deemed “false-positives”). Melting curve analysis revealed nearly all of these samples had atypical melting curve morphologies for the “Noro-1” target on BioFire (16/18, 89%), which was statistically significant (Odds Ratio 173.2, 95% CI [22.2, 5326.9],p<0.0001). Stool samples with multiple pathogens including norovirus detected by BioFire were not more likely to produce false-positive norovirus results (Odds Ratio 1, 95% CI [0.3, 3.3],p=1).ConclusionsAlthough not described in the manufacturer’sInstructions for Use, we propose routine manual review of melting curves for the BioFire GI panel prior to reporting, to avoid potential false-positive norovirus results.