Effects of universal masking on Massachusetts healthcare workers' COVID-19 incidence

Lan, F.-Y.; Christophi, C. A.; Buley, J.; Iliaki, E.; Bruno-Murtha, L. A.; Sayah, A. J.; Kales, S. N.

Background: Healthcare workers (HCWs) and other essential workers are at risk for occupational infection during the COVID-19 pandemic. Several infection control strategies have been implemented. Particularly, evidence shows that universal masking can mitigate COVID-19 infection, though existing research is limited by secular trend bias. Aims: To investigate the effect of hospital universal masking on COVID-19 incidence among HCWs compared to the general community population. Methods: We compared the 7-day averaged incidence rates between a Massachusetts (USA) healthcare system and Massachusetts residents statewide. The study period was from March 17 (the date of first incident case in the healthcare system) to May 6 (the date Massachusetts implemented public masking). The healthcare system implemented universal masking on March 26, we allotted a 5-day lag for effect onset, and peak COVID-19 incidence in Massachusetts was April 20. Thus, we categorized March 17-31 as the pre-intervention phase, April 1-20 the intervention phase, and April 21-May 6 the post-intervention phase. Temporal incidence trends (i.e. 7-day average slopes) were compared using standardized coefficients from linear regression models. Results: The standardized coefficients were similar between the healthcare system and the state in both the pre- and post-intervention phases. During the intervention phase, the healthcare system's epidemic slope became negative (standardized {beta}: -0.68, 95% CI: -1.06 to -0.31), while Massachusetts' slope remained positive (standardized {beta}: 0.99, 95% CI: 0.94 to 1.05). Conclusions: Universal masking at the healthcare system was associated with flattening the COVID-19 curve among HCWs, while the infection rate continued to rise in the surrounding community.