Rxivist logo

Stringent thresholds for SARS-CoV-2 IgG assays result in under-detection of cases reporting loss of taste/smell

By David W Eyre, Sheila F Lumley, Denise O'Donnell, Nicole Stoesser, Philippa C Matthews, Alison Howarth, Stephanie B Hatch, Brian D Marsden, Stuart Cox, Tim James, Richard Cornall, David I. Stuart, Gavin R. Screaton, Daniel Ebner, Derrick W Crook, Christopher P Conlon, Katie Jeffery, Timothy M Walker, Tim EA Peto

Posted 25 Jul 2020
medRxiv DOI: 10.1101/2020.07.21.20159038

Thresholds for SARS-CoV-2 antibody assays have typically been determined using samples from symptomatic, often hospitalised, patients. Assay performance following mild/asymptomatic infection is unclear. We assessed IgG responses in asymptomatic healthcare workers with a high pre-test probability of Covid-19, e.g. 807/9292(8.9%) reported loss of smell/taste. The proportion reporting anosmia/ageusia increased at antibody titres below diagnostic thresholds for both an in-house ELISA and the Abbott Architect chemiluminescent microparticle immunoassay (CMIA): 424/903(47%) reported anosmia/ageusia with a positive ELISA, 59/387(13.2%) with high-negative titres, and 324/7943(4.1%) with low-negative results. Adjusting for the proportion of staff reporting anosmia/ageusia suggests the sensitivity of both assays is lower than previously reported: Oxford ELISA 90.8% (95%CI 86.1-92.1%) and Abbott CMIA 80.9% (77.5-84.3%). However, the sensitivity may be lower if some anosmia/ageusia in those with low-negative titres is Covid-19-associated. Samples from individuals with mild/asymptomatic infection should be included in SARS-CoV-2 immunoassay evaluations. Reporting equivocal SARS-CoV-2 antibody results should be considered.

Download data

  • Downloaded 1,448 times
  • Download rankings, all-time:
    • Site-wide: 16,321
    • In infectious diseases: 1,859
  • Year to date:
    • Site-wide: 23,445
  • Since beginning of last month:
    • Site-wide: 49,333

Altmetric data

Downloads over time

Distribution of downloads per paper, site-wide