Officials aren’t sure exactly what’s driving the COVID-19 surge. Experts say that makes policy more difficult.

Since COVID-19 cases began to rise in late October, small gatherings were quickly pinpointed by state officials in Virginia and across the country as a significant contributor to the spread of the virus.

“We need to continue to figure out why those numbers are rising,” Gov. Ralph Northam said at a news conference in early November. “And they’re rising right now because people are gathering and they’re not wearing masks.” Three days later, he reinforced the message in a video announcing new statewide restrictions, including a midnight curfew for restaurants and a ban on gatherings of more than 25 people.

“We know this virus is spreading in indoor places like restaurants where people take off their masks,” he said in the video announcement. “It’s spreading at small social gatherings, like dinner parties, and it’s spreading when people ignore the science and think they don’t need to wear a mask inside.”

The new mandates were quickly criticized by some of the administration’s political opponents, especially when it came to the limit on gatherings sizes. “If this is to be believed then it seems [the governor] is looking to dictate how many people you can have in your own house for Thanksgiving,” Del. Kirk Cox, R-Colonial Heights — who announced his own campaign for governor last month— wrote on Twitter. House Minority Leader Todd Gilbert, a frequent critic of the governor’s pandemic emergency orders, called the new restrictions “drastic and confusing” in his own tweet following the announcement.

Incredibly disappointing to see the Governor announce drastic new and confusing COVID restrictions on his fellow Virginians late on a Friday afternoon, by video, and without taking questions. That is not how you govern.

— Delegate Todd Gilbert (@cToddGilbert) November 13, 2020

But the focus on small gatherings as a factor in the nationwide surge has also been increasingly questioned by many epidemiologists and health experts, some of whom have criticized policies that allow businesses such as restaurants to remain open while extended families are encouraged not to meet for the holidays. 

Virginia’s gathering restrictions are significantly more lenient than they are in states such as neighboring North Carolina, where indoor events are capped at 10 people, or Minnesota, where Gov. Tim Walz temporarily banned all indoor and outdoor gatherings for anyone who doesn’t share a household. But experts say it can still be confusing when public policies don’t seem to follow an understandable pattern — when family gatherings are discouraged, for instance, but indoor dining remains open.

“I think limitations on the size of gatherings does make sense to me,” said Tara Kirk Sell, a senior scholar at the Johns Hopkins Center for Health Security. “But I also think people expect a continuum of logic that goes between each restriction so they all make sense together. Otherwise, they’re asking, ‘Why this and not that?’”

Dealing with limited data

Underpinning the policy response is a more fundamental question: What’s really driving the spread of the virus? Local health departments across Virginia have reported outbreaks linked to specific gatherings and events, from a wedding in Richmond to a Thanksgiving-related exposure in Southwest Virginia that’s already forced multiple families to quarantine, Northam said at a news conference Wednesday.

But there’s currently not enough data to conclusively say what’s driving the increased spread across the state. Out of a total of Virginia’s 244,503 cases as of Thursday, 34,405 of them — about 14 percent — are connected to an outbreak, defined as two or more lab-confirmed cases linked to a single source, according to data from the state Department of Health. 

Roughly 52 percent of cases are associated with a known source, whether that’s an outbreak or another person who’s known to be infected.

“There’s a lot of anecdotal information,” said Elena Diskin, an epidemiology and COVID-19 containment program manager at VDH. “We have our daily huddles and meetings that bring together a lot of the information from the local health departments — the boots on the ground — and that’s probably our main way of gathering a feel for what’s going on in the state.”

But for nearly half of all cases, the cause of the infection isn’t always clear. Virginia’s case investigation script, used by public health workers to collect information from patients who test positive for the disease, does include questions on employment and whether patients have been reporting to the workplace. Case investigators also ask about other public-facing activities, from going out to eat and taking public transit to attending church.

A “yes” to any of those questions would warrant follow up, including more detailed interviewing about the level of risk in each situation. “If you said you had been out to a restaurant, I think the follow-up question is, ‘Can you tell me where you sat and what you did?’” said Rebekah Butterfield, an epidemiologist with the Richmond-Henrico Health District who’s been focusing on COVID-19. “‘Like, how close were the tables, were you masked when you weren’t eating?’ There are actually a bunch of questions to try to ascertain whether there were close contacts.”

The answers can paint a broader picture of factors that seem to be contributing to community spread. Dr. Sue Cantrell, director of the Lenowisco and Cumberland Plateau Health Districts in far Southwest Virginia, said Thanksgiving gatherings have been a prominent theme in case interviews since the holiday. Earlier in the fall, workplace exposures and a lack of mask use were common trends, as were gatherings, including “but not limited to: birthday parties, funerals, sports activities and religious services,” Cantrell wrote in a Thursday email.

“We have also noted an increase in people reporting a mild onset of symptoms (cold-like, headache or sinus infection) who do not think they have COVID and continue with regular activities,” she added. “[They’re] often only tested after persistence of symptoms or after others in their ‘bubble’ have symptoms and they are tested along with newly symptomatic contacts.”

But as COVID-19 numbers continue to surge, investigators are often limited by strained resources or a lack of information. And participating in multiple high-risk activities can make it more challenging to trace a case to a single source of exposure. 

Seth Levine, an epidemiology program manager for VDH, said there are sometimes lightbulb moments, such as several infected patients listing the same business or event as part of their daily activities. But just as often, a case might not list any particular high-risk exposures or decline to participate in the process altogether.

“Sometimes we have really awesome, thorough interviewees who give us all of the information they can think of,” Butterfield added. “They’ll share all the names and information of people they consider to be close contacts. And sometimes we’ll get someone who just gives us, ‘This is what I did, this is where I was, yes I’ve had close contacts, but I’m not going to share them with you.’”

VDH does share some data on contact tracing, though it’s not always reflective of how many Virginians participate in the interview process. Currently, the department reports that 72.5 percent of cases are contacted within 24 hours. Diskin and Butterfield said the number reflects when a local health department opens an investigation into a positive case, but doesn’t mean that the person has responded to calls or completed a questionnaire.

So far, the department has been unable to connect with nearly 12 percent of all case contacts across the state, which means it’s still unknown whether those Virginians were tested or went into quarantine to avoid potentially spreading the virus. Of known cases, Diskin said roughly 60 percent complete an interview with an investigator, based on VDH’s internal numbers.

“We did have to add some data fields to collect that information, so of course it’s only valid since we added those,” she said. And according to Butterfield, those interviews can yield significantly different levels of information depending on a person’s cooperation and how much they’re able to remember about their last two weeks of activity. 

Growing infections can be another challenge. Health officials in Roanoke have reported prioritizing cases on days when hundreds of new ones have come in, making it nearly impossible for investigators to reach out to the infected patient and to every close contact. Virginia has increased its public health workforce over the course of the pandemic, and not every district is in the same predicament — Richmond, for instance, is currently experiencing slow growth, while cases in much of Southwest Virginia have been surging for weeks.

But more infections can make it difficult for investigators to dive fully into the potential exposure history of every case. Butterfield said public health workers always aim to collect information about patients’ activities in the two weeks before they were officially diagnosed with the virus — as well as the two days before symptoms begin, when they’re most likely to have passed the disease to other people — to understand how they could have contracted the virus and might have spread it.

Often, patients remember more about that two-day period, which makes it easier for investigators to track the virus forward and try to mitigate potential spread. But even in districts with slower case growth, focus tends to turn on identifying cases with the highest risk of exposing others and then attempting to limit that spread.

“Under normal circumstances, the local health departments will try to investigate every single case patient and close contact,” Levine added. “But with the surge, that becomes very, very challenging. So, that’s where prioritization definitely becomes more important — identifying those high-yield cases or close contacts and targeting those individuals.”

‘It’s a tricky epidemiological question’

Experts say the limitations of existing case investigation data — along with an increasingly fatigued public health system that’s tasked with slowing an often-deadly and infectious disease — can make outbreaks linked to small gatherings both easier and more important to identify.  

“It’s a tricky epidemiological question,” said Bryan Lewis, a computational epidemiologist at the University of Virginia’s Biocomplexity Institute, which partners with the state health department to model the potential spread of COVID-19 across the state. “I don’t think there’s any reason to doubt that of the case investigations they’re able to conduct and identify, a large amount of the transmissions are from these smaller family gatherings.” 

“But those are what people tend to remember, and I imagine there’s a reasonable amount of the investigations where it’s just undetermined,” he added. “You’re more likely to be able to fill in that question mark if someone says, ‘Yeah, I hung out with these four or five people, they’re my friends and family, and here are their phone numbers.’ Then you might learn that two of them actually did get infected and you’re able to draw that link.”

There’s also universal agreement among experts that gatherings are a major risk factor when it comes to spreading COVID-19. While the virus can be transmitted through relatively brief interactions, the risk is much higher with sustained face-to-face contact in situations where one or more people aren’t wearing masks. Butterfield said that the biggest risk comes when multiple households spend an extended period of time together indoors — the exact conditions of most holiday gatherings.

In that sense, Sell said it’s logical for leaders to focus on small gatherings when they’re considering new restrictions. Northam’s spokeswoman, Alena Yarmosky, said the governor’s decision was informed at least partially by modelling from Virginia Commonwealth University’s Biocomplexity Institute, which shows the statistical risk of encountering a person infected with COVID-19 is “significantly lower at 25 people than it is at 250.”

“He also relied on the aggregated experiences of the state’s contact tracers and case investigators,” Yarmosky added in an email on Thursday. “While this data is certainly limited, health officials do report a significant number of outbreaks from social gatherings (identified as weddings, private parties, social clubs, or other get-togethers).”

Sell, and many other experts, said the ongoing pandemic placed state leaders, who have largely drafted policies with limited federal consensus or guidance, in a difficult situation. On one hand, apparent logical inconsistencies, like restricting gathering sizes while allowing non-essential businesses to remain open, could frustrate some residents and make them more skeptical of new restrictions. 

“Some people could be frustrated by any sort of internal discord, and that could make them question things,” Sell added. On the other hand, public health experts are concerned that COVID fatigue could make the remaining months of the pandemic worse, and any new sweeping restrictions — especially those that forced businesses to close — could make people even less likely to comply.

“We think this idea of more targeted interventions, now that we understand the disease better, is likely to work,” said Madhav Marathe, director of the UVA Biocomplexity Institute’s Network Systems Science and Advanced Computing Division. And experts largely agree that private gatherings can be riskier than restaurants or public-facing businesses like hair salons, where interactions are more likely to be distanced and owners are required to follow statewide health and safety guidelines.

Still, there’s no solid evidence that social gatherings, with similar safety protocols, are significantly more likely to contribute to the spread of COVID-19 than dining inside bars and restaurants, for example, which have been linked to outbreaks in Virginia and across the country. It’s also still unclear how the state’s most recent restrictions will affect case transmission. 

Early on in the pandemic, stay-at-home orders and other targeted restrictions had a significant impact on the number of people who chose to stay at home, said Aref Darzi, the project manager of the University of Maryland’s COVID-19 Impact Analysis Platform, which uses anonymized cell phone data to track mobility during the pandemic. But in recent months, those orders haven’t had the same impact on behavior. Marathe also said there’s less of a correlation between movement and the spread of COVID-19 in UVA’s modelling given that so many other guidelines can impact transmission — including the simple act of wearing a mask.

“Often, when we boil everything down, it still comes back to avoid large gatherings indoors and don’t go anywhere if you’re sick,” Butterfield, with the Richmond-Henrico Health District, added. It’s created a level of interpretability in the word “gathering,” which she said could range from an extended family get-together to indoor sports practices. But as health officials anticipate a continued surge in cases before a COVID-19 vaccine becomes widely available, they’re warning against taking any non-essential risk.

“I would recommend that anything that is avoidable should be given very serious consideration,” Marathe added. In the last nine months, he’s only been to a grocery store twice. Lewis, the UVA epidemiologist, said he hasn’t visited a restaurant since the beginning of the pandemic. 

“In terms of the risk mitigation, going to an indoor environment where there’s a reasonable number of people, even with mask wearing, should be considered a kind of risky endeavor,” he added. “For small family gatherings, it should be done very carefully. Or make the conscious decision to quarantine yourselves and then bubble up together. As we go into the winter, it just makes a lot of sense to make those decisions, because it’s going to take a few months for us to reach an end to this.”