Assessing the performance of regular surgical nose masks as a sampling method for SARS-CoV-2 detection in a cross-sectional study

Nose masks are widely worn for protection against respiratory pathogens, including SARS-CoV-2. They have been reported as possible substrates for viral sampling and testing for COVID-19 but, evaluations have so far been purposive; involving individuals known to have the infection and using improved materials on the nose masks to trap the virus. We investigated the feasibility of using the regular 3-ply surgical masks and, voluntary coughing as a mode of particle expulsion for detecting SARS-CoV-2 infections in a cross-sectional study at Ghana’s first COVID-19 testing reference laboratory, the Noguchi Memorial Institute for Medical Research, University of Ghana. Paired samples of naso-oropharyngeal swabs and nose masks already worn by 103 consenting adult participants (retro masks) were collected. Participants were also required to produce three strong coughs into a newly supplied sterile surgical nose mask. Pre-wetted swabs in Viral Transport Media (VTM) were used in swabbing the inner lining of each nose mask. The swabs used were then stored in VTM to maintain the integrity of the samples. PCR results of SARS-CoV-2 detection from the nose masks were compared to those from naso-oropharyngeal swabs (‘gold-standard’). Out of the 103 participants tested with all three methods, 66 individuals sampled with naso-oropharyngeal swabs were detected as positive, and the retro and new masks matched 9 and 4, respectively. Only 3 individuals were positive across all three sampling methods accessed. The retro nose masks performed better in matching the gold-standard results than the new mask + coughing method, with 90% vs 80% sensitivity, positive predictive value of 13.6% vs 6%, and a weak but significant linear relationship (adj. R(2) = 0.1; P = 0.0004). Importantly, we also show that the nose masks would work for sampling whether individuals are symptomatic or asymptomatic since gold-standard PCR cycling threshold (Ct) values for positive individuals did not differ between the two groups (P< 0.05). We recommend including features such as talking during participant engagement, use of a spontaneous cough inducer and increased coughing bouts > 3, to improve the performance of sterile nose masks for SARS-CoV-2 detection.