As the pandemic worsens in Europe and elsewhere, there is already talk about when public health officials in the United States should reimpose covid-19 measures such as mask mandates. But they should be wary about setting rigid “on-ramps” back to restrictions.
If the past two years have taught us anything, it’s that predetermined thresholds for public health policy don’t work well in practice and risk undermining the public’s trust.
Time and again, the fast-moving pandemic has shown us that predictions age poorly. Scientific knowledge evolves, and metrics set at one floating moment have often proved ineffective to guide policymaking in the real world.
Experts and policymakers have tried countless combinations of metrics to guide decisions, from case numbers to test positivity to community-level vaccination rates. In theory, such metrics can keep transmission to a minimum, prevent outbreaks of severe disease and, if vaccination rates are included, encourage vaccination uptake. In the absence of a clear alternative strategy for lifting mandatory masking policies, we, too, encouraged policymakers to implement “off-ramp” metrics in the summer and fall.
But in reality, thresholds to change covid policies are always somewhat arbitrary, and the context is constantly changing. Because of the time preset measures are triggered, they are often out of date.
Last summer, for example, Massachusetts set a simple off-ramp for mask mandates in schools: If a school had an 80 percent vaccination rate in the building, then its leadership could apply for a mask waiver. We supported this plan, arguing that a high vaccination rate would provide strong protection against outbreaks of severe disease in schools, even in the face of the delta variant.
But few schools applied for the waiver, and even fewer lifted the mandate. why? Because some local health and district leaders opposed it, arguing that because the Pfizer/BioNTech vaccine for 5-to-11-year-olds would soon be available when the metrics went live, they felt they should wait “just a bit longer.” Then omicron arrived, and the vaccine was no longer as effective for preventing infections (although it was still highly protective against severe disease). In other words, the context of the pandemic changed.
A similar thing happened in many schools during the 2020-2021 school year, but in reverse. Schools established thresholds, such as rates of community spread, that would trigger the transition back to remote learning. Many of them crossed those thresholds during the winter peak, but by then, the situation had evolved. In-school transmission proved rare enough that following the pre-specified, arbitrary plans set over the summer did not make sense.
Using today’s metrics to guide school policy in tomorrow’s surges would be a mistake. Scientific factors interact in complex and unpredictable ways. Perhaps new advancements, such as better or more accessible treatments for those who become sick, will dramatically decrease community risk. Or perhaps the next variant will be more evasive of previous immunity or resist therapies. Any of these factors would complicate the real-world application of a pre-specified on-ramp.
We support the Centers for Disease Control and Prevention’s conceptual shift to focus its latest framework for determining covid policies on cases of severe disease and hospital capacity. But when it comes to school mask policy, what is happening in hospitals might have little relationship to what is happening in schools. Local factors — such as vaccinations, prior infection, who is being hospitalized and whether admissions are “for” covid — should all be considered in determining the need for public health interventions.
So what can be done to plot out future strategies? First, public health authorities should incorporate up-to-the-minute scientific information, while being transparent about what is uncertain and acknowledging that things might rapidly change. This will require further investment in the public health infrastructure to ensure robust, real-time surveillance and data analysis.
Second, decision-makers should demand continuous reassessment of policies that acknowledge past successes and failures. They should also opt for strategies tailored to local situations, rather than one-size-fits-all standards, and work to collect scientific evidence to support the effectiveness of metrics for achieving a policy goal.
Finally, we should set clear goals of any intervention and update those goals along with metrics in real time. This will improve our understanding for how, when and for how long interventions will be recommended.
It’s time to move away from default, static metrics. The public is remarkably adept, especially in the information age, at sensing when policy decisions are arbitrary. Americans can tell when rules lack a clear evidence base or are not relevant to the specifics of the situation in any given community. Accepting uncertainty and rapid change is scary, but it is also our only realistic path forward.