Tensions are high right now. As the delta variant spreads like wildfire across the U.S., vaccination rates are still low in many places and parents and school staff are anxiously wondering what will happen when schools start up again. Should there be more mask mandates? Will businesses have to close again? Will big gatherings be banned?
There are no clear, simple answers to these difficult questions — in some places, officials and employers in the same locale are making different decisions. But for those looking to the scientific evidence for guidance, here are some things a team of researchers at the University of Pittsburgh and Carnegie Mellon University have learned since the pandemic began that can help.
The overall message from their analysis of state-by-state U.S. data is this: Even moderate restrictions on the size of gatherings, and some level of mask requirements can help help slow a surge of COVID-19 cases, if implemented at the right time.
The scientists have kept a detailed database of statewide restrictions from March 2020 to March 2021, documenting when they went into effect, when they were lifted and the impact they had on case and death numbers in each state. They recently published a preliminary analysis of that data online, which is in the process of being peer-reviewed.
"As we go into the fall, when schools are going to go back in person and we do not have the vaccine coverage that we need to have to reduce the spread of SARS-CoV2, we need to be thinking harder about putting interventions on," says Seema Lakdawala, an assistant professor and flu researcher in the department of microbiology and molecular genetics at the University of Pittsburgh who was one of the project leads.
"It doesn't mean that we have to stay at home and we can't go out anymore and we can't see our friends anymore," she says. It just might mean bringing back some level of masking requirements and gathering restrictions to help keep transmission under control.
As communities and policymakers consider how to manage the threat of this delta surge, here are some lessons to keep in mind that Lakdawala's team and their collaborators at Carnegie Mellon University have learned from the first 18 months of pandemic restrictions.
Throughout the pandemic, states — and even counties and cities within those states — have often done their own thing. Transmission has varied; residents' willingness to comply with restrictions has varied. Especially in the beginning, it was a real grab bag, says Lakdawala.
"A lot of states were putting [restrictions] on and then taking them off and then putting them on and taking them off — it was really haphazard," she says.
To understand whether — and how — each type of restriction was making an impact on case trajectories, Lakdawala's team divided them into categories — stay-at-home orders, nonessential business restrictions, indoor gathering limitations, restaurant or bar restrictions, and mask mandates — and assessed how stringent each of the measures were.
Then, with the help of Rebecca Nugent, the head of the Department of Statistics and Data Science at Carnegie Mellon University, they created graphs of each state's cases and deaths, showing the dates when restrictions were put into place or lifted.
The first clear takeaway from the analysis of all this data from all these states, Nugent says, "is that these types of restrictions work — that they have an impact, that they are contributing to the control of the spread, and that we should be holding them on probably a little bit longer than we think."
The timing of interventions is really important, Lakdawala says. The analysis found it was better to put restrictions in early — before a surge has really had a chance to heat up.
"Things are not instantaneous," she says. It wasn't until about three or four weeks out that the case and death trajectories started to respond to the restrictions.
Sometimes, too, it can be too late for the restriction to turn a surge around. If cases are already spiking, "sometimes it's already past the point of no return." Many people may already be incubating the virus, just not yet sick, by the time restrictions are put into place.
"That's why we keep advocating for putting them on earlier so it can be a smaller surge for a shorter amount of time," Nugent explains. "That's what we're aiming for."
Bottom line from the data: "Oftentimes, we're waiting a little bit too long and we're taking them off a little bit too early," says Nugent. It pays off to get a jump on restrictions before cases have really started to accelerate and to wait at least four weeks before lifting them.
In analyzing states' total numbers of COVID-19 cases over time at various levels of restrictions, the researchers found there seemed to be a sweet spot: Places that had several kinds of moderate restrictions in combinations did really well in damping down viral spread.
Perhaps surprisingly, "states that tended to have [more stringent] interventions weren't necessarily doing drastically better than those with moderate interventions," Lakdawala explains.
What seemed to work just right — what they called the "Goldilocks" combination — was "some level of masking, some level of restaurant or bar restrictions and some level of gathering-size bans," she says. "Those — we think — are the critical three."
"Once that happened and it was held on long enough, that's when we saw an impact on the number of cases [and] the number of deaths," Nugent adds.
Melissa McPheeters, an epidemiologist at the independent research firm RTI International, says it makes sense that this team found that the interventions work best in combination.
"It gets back to that Swiss cheese analogy that lots of people use," she says. None of the interventions are perfect, "so you layer them to ensure that you're putting as many possible barriers as you can in between people and the virus."
Minnesota and South Dakota had vastly different approaches to statewide restrictions. On a 5-point scale, where 5 is most restrictive, South Dakota's median score was 0 and Minnesota's was around 2.5.
But when the researchers ran calculations to see which states had similar case curves and put those results on a map, the two side-by-side states were quite similar. That swath of states with a similar curve ran all the way from Montana to Wisconsin in the Upper Midwest. Another clear region in the South stretched from Texas to Virginia, and another ran from West Virginia up the East coast to New Hampshire.
"I was just floored when I saw the map," Lakdawala says. "It was clearly geographically clustered in a way that you wouldn't expect when you look at just states and their interventions."
That clustering "makes sense if you think about it," Nugent says. In each region there may be "similar weather patterns, people traveling across state lines," she says. Weather and season can affect viral spread and transmission dynamics, Lakdawala says.
"I think as we move forward, we really need to think about that, and having states talk to each other a little bit more [and] to do interventions at the same time to help reduce burdens across their states," Lakdawala says. These findings suggest that states can't just make changes within their borders and expect to control their fates.
"Clearly we're much more connected than anyone really thinks about in their day-to-day lives," says Crystal Watson, a senior associate and assistant professor in environmental health and engineering at the Johns Hopkins Center for Health Security. "We need to think about how we make more unified choices about how to approach infectious disease emergencies as a country."
One question this research couldn't answer is whether some restrictions are better than others — as in, are restaurant capacity caps more important than limiting the size of other gatherings?
"We don't have that, because clusters of interventions were implemented routinely and then eased routinely," Lakdawala explains. "So it is really hard to tease it apart."
The scientists also only analyzed the data at the state level, even though a lot of states left it to counties or even cities and towns to come up with their own restrictions. Lakdawala says her team is next working on an analysis of whether some counties did better than others.
The cultural context of a place matters too, McPheeters notes. Even without statewide action, people sometimes change their own behavior in response to hearing about rising case numbers where they live and hospitals filling up. "But the degree of that change — and whether or not that happens — probably is different based on the social context of a particular area," McPheeters says. "So if you have a lot of people in an area who believe that masks work, versus an area where people don't believe that they work, you're going to have different kinds of responsiveness to what's going on in terms of cases."
Watson says more detailed analyses could answer some key outstanding questions. "If I were a policymaker, I would want to know what are the thresholds at which you need to put in place these different layers of mitigation?" she says. "What is the stringency of each of these measures, what is the duration of how long they should be applied, and what is the optimal combination of mitigation measures for the outbreak that you're in?"
She says she thinks this research is "a good start to a much more in-depth investigation that we need — over time — to really understand how these mitigation measures affected outcomes and what actions we should take for the next infectious disease emergency."
Although this research focused on mandates and other restrictions to control the spread, COVID-19 vaccines are an incredibly important part of that picture, too. What's important for now, these scientists say, is to use these other interventions to help keep down the virus' spread until more people can get vaccinated.
The delta variant is indeed highly contagious. But the good news is that all of these tools — including wearing a good mask where appropriate, keeping gatherings small, washing your hands frequently and keeping fresh air circulating indoors — can still work to help stop the deadly virus in its tracks.