Among medical care workers who offered routine care to individuals infected with COVID-19, the overall estimates ruled out a doubling in the hazard ratios (HRs) of RT-PCR–verified COVID-19 for medical masks in comparison with HRs of RT-PCR-verified COVID-19 for N95 respirators, according to a study published in Annals of Internal Medicine.

It's unclear whether medical masks and N95 respirators offer comparable protection against coronavirus disease. To ascertain if N95 respirators are superior to medical masks for regular care providers to evade COVID-19, a randomized, noninferiority, open-label trial was carried out. A total of 1009 medical personnel were included who offered direct care to patients having suspected or confirmed COVID-19. The policy implemented at each site was the usage of medical masks for 10 weeks instead of fit-tested N95 respirators, along with universal masking.

Confirmed COVID-19 on RT-PCR test was the key endpoint ascertained. In the intention-to-treat assessment, volunteers in the medical mask group had RT-PCR-confirmed COVID-19 in 52 of 497 (10.46%) cases in comparison with 47 of 507 (9.27%) cases in the N95 respirator group (HR, 1.14). Unplanned subgroup assessment by country revealed that RT-PCR-substantiated COVID-19 occurred in 8 of 131 cases (6.11%) in the medical mask group vs 3 of 135 cases (2.22%) in the N95 respirator group in Canada (HR, 2.83), 6 of 17 cases (35.29%) vs 4 of 17 cases (23.53%) in Israel (HR, 1.54), 35 of 257 (13.62%) vs 38 of 261 (14.56%) in Egypt (HR, 0.95), and 3 of 92 (3.26%) vs 2 of 94 (2.13%) in Pakistan (HR, 1.50).

In the medical mask group, there were 47 (10.8%) adverse events connected to the intervention, compared to 59 (13.6%) in the N95 respirator group. The aggregate estimates ruled out a doubling in risk of RT-PCR-substantiated coronavirus disease for medical masks in comparison with HRs of RT-PCR-verified COVID-19 for the N95 respirators among healthcare professionals who regularly offered treatment to coronavirus-infected patients. Owing to treatment effect variability, the subgroup results varied by country, therefore the aggregate estimates might not be applicable to specific nations.