For regions shut down due to Covid-19 to safely begin to reopen, we need ways to keep R — the average number of additional people infected by each infected person — under one, the threshold below which epidemics contract and ultimately die out. Among the proposals for how we can do this in the United States, one calls for frequent population-wide testing to identify and isolate those who are infected. Others suggest that the country will be hard-pressed to get through this without either prolonging lockdowns or intermittently reinstating them whenever infections rise until we have enough testing and contact tracing to control the spread or enough people become immune through infection or vaccination. The former requires testing on a scale that, barring a breakthrough, will not be possible anytime soon. The latter would inflict ongoing social and economic damage with the specter of lockdowns constantly looming over us.
We suggest another way that is perhaps both more achievable in the near term and sustainable over the long term. It is based on our belief that given the expected limited levels of testing that will be available for the next several months, we may not be able to detect and isolate enough infected people to keep R below one without lockdowns even with strong contact tracing.
Once cases are declining for multiple weeks, though, we could begin easing social distancing if we can implement population-wide social protections that, combined with more modest increases in testing and tracing, could be sufficient for keeping R below one. Social protections are ways to protect against transmission, including asymptomatic and presymptomatic spread, that still allow people to work and resume some normal activities. Our plan involves two measures:
Mass producing and then widely distributing masks that are more protective than those that are now typically being worn by the general population
Ensuring rigorous implementation of physical distancing and hygiene in workplaces, public areas, high-risk settings (e.g., homeless shelters), and homes
While many reopening plans call for versions of these measures, we are calling for governments, employers, and the population at large to be bolder, more systematic, and more innovative in maximizing these approaches. In the absence of adequate testing, strong social protection is the only other lever we have to blunt transmission enough that lockdowns could potentially be relaxed without needing to be quickly reimposed.
Ultimately, social protection may neither need to be as effective as social distancing nor require testing and tracing to be perfect. The key is for their combined effect to be good enough to keep R below one. This could be more achievable than it seems. From an analysis we conducted (not yet peer reviewed), if social protection were 50% effective, we would keep R below one by isolating 40% of symptomatic infected individuals within a day of the onset of symptoms.
Invisible Spread
One of the key reasons why current levels of testing, tracing, and isolation are not enough to stop Covid-19 is its “invisible” transmission. Between 25% and 80% of infected people have no symptoms, or only mild ones, yet still infect others, some possibly contributing to “superspreading” events. Even patients who develop serious illness may be most infectious either one or two days before their symptoms start or on the day that their symptoms appear. Almost half of all transmission may happen during this presymptomatic period when people — and those around them — don’t know they are transmitting.
The White House projects that five to 10 people need to be tested to find one infected person; others suggest that number is 50 to 100 people. At current levels of infection, testing widely enough to stop enough asymptomatic and presymptomatic transmission would likely require millions of tests per day — more than the roughly 200,000 a day done now or the 450,000 a day expected by the fall. Contact tracing can find some of these invisible transmission chains, but no matter how many tracers are hired or what digital tools are used, it can only help to the extent that testing is available to identify cases whose contacts need to be traced, isolated, and also tested.
Social Protection Strategies
Many states have recently gotten R just below one by buttressing testing with lockdowns. As states gradually reopen, we need measures that mimic the protective benefits of lockdowns without their destructive downsides.
High-filtration masks. Though Covid-19 can spread through surfaces and contact, it seems to mainly transmit through the air. If we block this respiratory transmission, we should be able to control the virus. High-filtration surgical masks that are easier to wear than N95 masks can help achieve this goal. They could be just as important to stopping Covid-19 as any diagnostic or treatment.
The cloth masks that people are now using vary widely in how well they block infection but typically stop less than 50% of viral particles with many closer to 20% or less. High-filtration surgical and N95 masks used by health workers more reliably impede transmission. Right now, these are rightly prioritized for health workers. What is more, N95 masks are difficult to wear for long periods of time even for those used to them.
While not as protective, high-filtration surgical masks are generally more effective than cloth masks and more wearable than N95s. A new study suggests that combining cotton with other widely available materials, such as silk, chiffon, or flannel, could achieve levels of filtration similar to these masks.
An existing high-filtration surgical mask or a new design — ideally one that is reusable — that strikes the right balance between protection and comfort should be rapidly mass manufactured and distributed to the general population. As was done for ventilators, we need to use the Defense Production Act to mass produce and widely distribute these masks.
While this sounds ambitious, it may be easier and faster to do than establishing adequate testing. And, though there is a concern that wearing masks may prompt people to be less careful in other ways, we have not seen any data to support this notion. In fact, similar reservations were raised about whether seatbelts would cause careless driving, which studies have shown to be untrue.
Physical distancing. The Centers for Disease Control and Prevention (CDC) needs to develop clear guidelines and regulations for maintaining safe distancing in public that local health departments can use to help businesses implement and then monitor for compliance. Even once lockdowns are relaxed, we need to keep large public gatherings on hold. Public areas need to be choreographed to ensure spacing — for instance, by limiting the number of passengers in a subway car or customers in a business at any one time. Workplaces similarly need to be reorganized to minimize crowding — by staggering shifts, limiting in-person meetings, spacing out seating arrangements, resorting to telework as much as possible, and so on. Restaurants and retail stores need to actively plan and manage the spacing of customers, provide hand sanitizing facilities, and ensure appropriate ventilation to prevent viral particles from lingering in the air.
Hygiene. While masks and distancing address respiratory transmission, fomite spread — spreading a disease through surfaces — needs to also be blocked by routinely disinfecting highly frequented areas and making hand sanitizing ubiquitous in public spaces. While preventing people from picking up virus this way, we also need to other strategies to nudge them into not touching their face and mouth. In a study observing medical students, subjects touched their face 23 times per hour. While changing these behaviors is difficult, it’s not impossible. For example, wearing rubber or other reusable gloves when grocery shopping may make people less likely to touch their face. Other creative, scalable, and possibly simple ideas could go a long way.
Home isolation. People with symptoms, confirmed infection, or identified as a contact of an infected person need to be isolated until they are clearly not infectious. If isolating at home — where household members are up to 20 times more likely to get infected than other contacts — people need to be truly isolated. They should not share bathrooms, beds, or living spaces with others and should wear masks and wash hands before passing through common areas. People for whom this is not practical — for example, those living in crowded housing or with people who are at high-risk — need to be given the option to isolate in hotels, dorms, or other repurposed venues free of charge.
While people seem willing to adopt protective practices at home when someone is symptomatic, it may be unrealistic to expect people to do so when family members could unknowingly be asymptomatically or presymptomatically infectious. Scenarios may arise where guarding against this invisible spread by wearing masks and practicing social distancing within households might become important for controlling the epidemic, especially as people return to work. However, such intrusive measures would have to be weighed carefully against their extreme social cost. The number of people living in a household would likely be an important factor in navigating such situations. For community housing scenarios — like nursing homes and homeless shelters, where large numbers of people get infected very quickly — the need for protective measures is clearer.
As states look to reopen, we need to establish a multilayered, social-protection strategy that, combined with more achievable levels of testing and tracing, could keep R below one. Doing so will also require cultivating public buy-in without regressive punitive enforcement while supporting disadvantaged communities to adopt these approaches. We need to move quickly to create and widely implement such a strategy within the coming weeks, not months.