Discussion

In 10-minute speaking, SARS-CoV-2 was detected on the inner surface of the N95 respirators and surgical masks, but the outer surfaces of the respirators and masks were not contaminated by the virus. N95 respirators decreased the arrival of SARS-CoV-2 at a distance of 30 cm compared with no-mask wearing. Taken together, N95 respirators blocked the emission of SARS-CoV-2-containing particles. However, SARS-CoV-2 was identified in the air even when patients wore N95 respirators and surgical masks. Therefore, our findings suggest that universal N95 respirator or mask wearing by infected patients for source control and uninfected persons for inhalation protection is important to prevent COVID-19 transmission via the air in inadequately ventilated spaces.

We found that SARS-CoV-2 was detected in the air even when N95 respirators and surgical masks were used, especially in patients with high viral loads. There are several possible explanations for the results. First, preexisting SARS-CoV-2 might have been floating in the air of the patients’ rooms, which were inadequately ventilated, before the mask wearing. Bushmaker et al. showed that SARS-CoV-2 Omicron variant were durable in the air with a half-life of 2.15 hours. ^{Reference Bushmaker, Yinda, Morris, Holbrook, Gamble and Adney5} Second, suboptimal fitting of N95 respirators and outward leakage of exhaled air might be associated with air contamination. Similarly, a previous study reported that even N95 respirators did not completely block emission of SARS-CoV-2-containing aerosols in laboratory setting. ^{Reference Ueki, Furusawa, Iwatsuki-Horimoto, Imai, Kabata and Nishimura6} Third, possible penetration of virus-containing aerosols through surgical masks might contribute to air contamination. ^{Reference Kim, Bae, Kim, Park, Lim and Sung4}

The Omicron variant of SARS-CoV-2 is thought to have higher transmissibility via the air than prior variants. ^{Reference Lai, Coleman, Tai, German, Hong and Albert7,Reference Ong, de Man, Verhagen, Doejaaren, Dallinga and Alibux8} It might be attributable to the Omicron variant having greater tissue tropism for conducting airways than lung parenchyma compared to previous variants. ^{Reference Hui, Ho, Cheung, Ng, Ching and Lai9} High concentration of Omicron variant in conducting airways might be likely to generate SARS-CoV-2-containing aerosols in daily life via bubble film burst mechanism. ^{Reference Lai, Coleman, Tai, German, Hong and Albert7} We tried to identify the risk factors of air contamination by Omicron variant in the study. In consistent with previous studies, ^{Reference Lai, Coleman, Tai, German, Hong and Albert7,Reference Ong, de Man, Verhagen, Doejaaren, Dallinga and Alibux8} the risk of air contamination was associated with a high nasopharyngeal viral load of patients. However, Lai et al. reported that saliva viral load showed stronger association with air contamination than nasopharyngeal viral load, ^{Reference Lai, Coleman, Tai, German, Hong and Albert7} whereas saliva viral load was not different between patients with and without air contamination in our study. The discrepancy might be due to absence of standardized procedures for saliva collection, heterogenous processing of saliva, and reduced sensitivity of PCR assay in saliva sample. ^{Reference Tan, Allicock, Armstrong-Hough and Wyllie10} In clinical characteristics, there were no discernible differences between patients with and without air contamination. The unpredictability of air contamination also highlights the importance of universal masking.

Our study had several limitations. First, the small number of patients makes it difficult to draw hard conclusions. However, given the limited data regarding the spread of SARS-CoV-2 in real-life patients with COVID-19 wearing masks, our findings could be helpful in understanding COVID-19 transmission via the air. Second, baseline and downtime air samplings were not performed. Preexisting SARS-CoV-2 and carry-over effects that lead to SARS-CoV-2 accumulation in the air over time might influence the results. Third, we did not measure the quality of the speaking such as loudness that might affect the virus release. Finally, we did not exam viability of SARS-CoV-2. However, we tried to enroll patients in early periods from symptom onset who are likely to shed viable virus.

In conclusion, universal masking in infected and uninfected individuals is important to prevent COVID-19 transmission via the air in inadequately ventilated environments.