The Centers for Disease Control and Prevention (CDC) introduced a three-level system in late February to describe COVID-19 community impact in the more than 3,200 counties across the country. Aside from recommendations on vaccines and boosters, as well as testing if symptoms are present, these levels are being used to assess the need for face masks when indoors in public venues.
Almost four months later, it has become apparent that their guidelines for individuals are both inappropriate and misguided.
The low level has no masking recommendations. The medium level recommends that those at high risk of severe disease to consult their health care provider about the need for masking and taking other precautions. The high level recommends masking indoors when around other people in public venues.
The levels are based on three metrics: A county’s number of new COVID-19 cases and number of new COVID-19 hospital admissions, both scaled as rates (per 100 thousand population), and the percentage of hospital in-patient beds occupied by COVID-19 infected patients.
If the new cases rate falls below 200, then the community level is low unless the other two metrics reach or exceed certain thresholds. Once the new cases rate reaches 200, a county level jumps to medium. The high level is only reached once the new COVID-19 hospital admission rate, or the percentage of hospital in-patient beds occupied by COVID-19 infected patients, exceed a specified threshold.
The three metrics are embedded in a tiered evaluation system, akin to using if-then-else rules in computer programming. By design, the levels are suited to inform hospitals about impending risks and the necessity to prepare for possible increased admissions.
The levels do not capture individual risk since the driver for the high level are hospital admissions and in-patient bed capacity. Thus, they are inappropriate to guide individual decisions regarding mask use.
Another issue is that the three metrics are capturing where a county has been, not where a county is heading.
On April 15, the CDC reported 14 high-level counties, with the seven-day average of new cases at 36,000 cases.
Fast forward to June 10, there are 314 high-level counties, the most ever reported, with the seven-day average of new cases now around 110,000, over three times higher than the number reported April 15. Moreover, on April 15, over 94 percent of counties were at the low level. This percent has declined every week since then, now standing below 58 percent.
The CDC community levels provide useful information to assess the health of a county’s hospitals, but the CDC has neglected the prevention part of their responsibilities. Their masking advice is far too opaque to help individuals, essentially delegating prevention to the back burner.
Disaggregating the CDC community levels from masking guidance is the best way to allow them to provide value.
So what advice is needed for when to wear face masks?
Focusing solely on the number of new cases (the first metric) reaching a threshold is sufficient to inform people that the virus is actively spreading in a county. But what is the right threshold?
Many people are now using rapid in-home tests, with positive results not reported to the CDC national database. This means that a baseline of 200 cases per 100,000 population is 200 cases reported to the CDC per 100,000 population. The actual number of cases are estimated to be five to 10 times higher, which means that the 200 threshold should be lowered to provide meaningful information, perhaps as low as between 20 and 40.
As the virus continues to circulate and case rates remain substantial in an area, masking will be beneficial to further suppress the spread of the virus and reduce transmission risk. High-quality face masks (like N95 respirators) reduce the risk of transmission, provided they are worn correctly.
If the CDC community levels continue to guide masking, people will be misled down a rabbit hole that serves no one’s best interest.
We can all agree that mask mandates are not the solution. What we need are sensible masking advice that is easy to understand and credible.
Sheldon H. Jacobson, Ph.D., is a professor in Computer Science and the Carle Illinois College of Medicine at the University of Illinois Urbana-Champaign. A data scientist, he applies his expertise in data-driven risk-based decision-making to evaluate and inform public policy.
Janet A. Jokela, MD, MPH, is the acting regional dean of the University of Illinois College of Medicine at Urbana-Champaign. She is an infectious disease and public health physician.