In the last weeks of 2022, many counties crossed the Centers for Disease Control and Prevention (CDC) threshold from low and medium risk to high risk, triggering a return to mask mandates, particularly in educational settings. The goal of any standardized risk framework, such as the CDC “community risk levels” is to compare “apples-to-apples.” However, as has so often been the case throughout the pandemic, different state and local policies mean that the current COVID-19 county risk framework is comparing apples to oranges — and these faulty comparisons are impacting local policy responses and recommendations.
In February 2022, the CDC released its updated — and controversial — risk classification scheme for evaluating community COVID-19 burden. Updated weekly, the CDC’s “community risk level” map categorizes each county in the U.S. as “low risk” (green), “medium risk” (yellow) or “high risk” (red). These categories are associated with different public health recommendations; most notably about when to wear masks. According to this guidance, when a county is at low risk masks are not recommended, at medium risk masks are recommended for those at high risk of severe COVID-19, and during high-risk periods masks are universally recommended in indoor spaces. Unlike its prior risk-level map, the now one-year-old approach takes into account not only the number of new COVID-19 cases per capita but also new admissions to the hospital and the percent of inpatient beds occupied by COVID-19.
Today, it is well known that among those patients classified as “COVID hospitalizations,” some are admitted “for COVID” (as a result of COVID-19) and some “with COVID” (for other reasons but with a positive COVID-19 test). Only in Massachusetts has there been an attempt to quantify this. Since January 2022, every hospital in the state has been reporting the number of patients hospitalized with COVID-19 who have and have not received the drug dexamethasone, which is the standard of care for COVID-19 with lung involvement. Importantly, that proportion has been changing over time. A year ago, about half of patients had not been treated for COVID-19. In the last several months, that number is consistently around 70 percent. This means that not only are the “COVID hospitalizations” included in the CDC numbers not truly representative of the risk they are meant to measure, but the county risk levels have lost their meaning as preventative and therapeutic options have improved.
The U.S. has had a fragmented approach to COVID-19. One manifestation of this is that some states and some hospitals, have elected to test every patient upon admission for COVID-19 regardless of symptoms, while others never did. Still others took this approach earlier in the pandemic but phased the policy out at some point. This variability in practice affects the county level risk classification, as hospitals which do asymptomatic testing find more cases than those that do not.
Yet even though we know the metrics are measuring different things, depending upon where you are and the state in which you live, they are being used to dictate local public health policy, particularly with respect to masking in schools. There is almost certainly high correlation between counties or states that have more in-hospital testing (and therefore artificially inflated estimates of community “COVID risk level”) and those that tend to strongly recommend or mandate masks in response to CDC data. But is it fair for children in Massachusetts, Michigan or California to be subjected to a mask mandate simply because their local hospitals do more asymptomatic testing?
In December, the Society for Healthcare Epidemiology issued a strong statement discouraging routine asymptomatic testing for COVID-19 upon hospital admission, citing the potential for unintended harms; their policy position was subsequently supported by other national health care organizations, including the American Society of Anesthesiologists which has taken a position against pre-operative testing. In response, we can expect to see more and more hospitals discontinue universal screening of all patients over the next weeks to months — which will also impact apparent “risk level” according to current CDC risk-level map.
A theme of the pandemic is that the U.S. public health infrastructure needs to adapt — and quickly — to changing conditions on the ground. Metrics that made sense based on the low levels of immunity and limited treatments available in 2021 need to be updated. As the epidemiology of COVID-19 has changed, hospitalization metrics need to be adjusted to account for changing disease severity and testing policies. The metric adopted by the state of Massachusetts is a good example of a real-world policy implementation that works and is easily scalable.
The CDC risk-level map is simultaneously “not enough” for some and “too much” for others. If you are an individual at an increased risk for severe illness with COVID-19., a risk-level map that depends on the number of COVID-19 patients in the hospital is too little, too late: You should care about other respiratory pathogens in addition to COVID-19, and you want to know the risk is elevated before the hospitals start feeling the strain. Metrics that focus on early indicators, such as wastewater data or increases in testing positivity rate that occur early in a surge are more appropriate as a “weather map” for high-risk individuals. But if you’re someone with low risk, an inaccurate metrics framework that means different things depending upon where you live is not helpful for guiding local policy decisions. It is time to update the metrics again, to meet the needs of where we are, not where we were.
Shira Doron, MD, is the chief infection control officer at Tufts Medicine health system and an infectious disease physician and the Hospital Epidemiologist at Tufts Medical Center. She is an associate professor of medicine at Tufts University School of Medicine. She has advised the Massachusetts governor and commissioner of education.
Westyn Branch-Elliman, MD, MMSc, FSHEA, is an associate professor of medicine at Harvard Medical School and an infectious diseases specialist. She is also an associate editor at Infection Control and Hospital Epidemiology. She has advised the Massachusetts Department of Elementary and Secondary Education.
The views expressed are their own and do not necessarily reflect those of affiliated organizations.