As the Covid-19 Delta variant shifts what we consider to be “safe” and “unsafe” activities, a new online tool called the microCOVID Project is attempting to quantify those evolving risks. The model assesses a given scenario—say, a vaccinated person eating at an indoor restaurant with friends—in a specified county of the United States and produces an estimated risk of contracting Covid-19.

“It gives you a kind of common playing ground for understanding all sorts of activities,” says Ben Shaya, a contributor to the microCOVID project. “It separates some of the emotional load out of it.”

The project started after Shaya’s “pandemic pod” of friends and housemates wanted to be more equitable about safety precautions, so they took a scientific approach to their decision-making. “They wanted to have a better way to be fair about what people could do,” he says.

The group of friends created a mathematical model that draws on the latest research on masks, vaccine efficacy, current cases in each county and more, before translating that into a numerical value. They started the project in May of 2020 for their own use, and within a few months, created a version for the public. The site includes two tools: a calculator , and a risk tracker . The calculator is designed to give users an idea about the risk of a single event or scenario, while the risk tracker tallies a person’s activities to create an overall risk score that can be shared with others. The team that started as a handful of friends incorporated dozens of volunteer collaborators over the next few months including mathematicians, data scientists and a primary care clinician.

The group decided to measure risk in a unit they named microCOVIDs: a one-in-one-million chance of becoming infected with Covid-19. The website defaults to budget 10,000 microCOVIDs per person per year, which works out to a 1 percent chance of contracting the virus, according to the site.

To use the calculator, users select different scenarios, like attending a crowded party or meeting a friend for dinner, and enter details about what mask they’re wearing, if they’re fully vaccinated, if they’re indoors or outdoors. The estimation jumps between color-coded scales of riskiness. For example, a fully-vaccinated person in Manhattan wearing an N95 mask might have a risk of around 4 microCOVIDs if they went grocery shopping for one hour. Within the model’s default budget of 200 microCOVIDs per week, this is considered a “low risk” activity. Another Manhattanite in the same scenario, but who is unvaccinated and unmasked, would have a risk closer to 80 microCOVIDs—nearly half of the weekly allotment.

Users can also build a custom scenario, like attending an outdoor wedding in Miami. Let’s say the person is unvaccinated, and there will be around 100 guests of varying vaccination status. Because people will be eating and drinking, no one will be wearing a mask, and people are likely to be spaced at a normal social distance for about five hours. In this scenario, the user’s risk is around 3,000 microCOVIDs, or “dangerously high risk.” If all other factors stay the same, but that person is fully vaccinated, their estimated risk of infection falls to a “very high risk” of 500 microCOVIDs.

The site’s risk tracker and calculator are adaptable, so people can change the budget depending on their concerns and the vulnerabilities of people around them. “Nobody thinks they're being dangerous,” says Jenny Wong, a contributor to the microCOVID project. “Careful means different things with different people.” The risk tracker, which was the original form of the project, works more like a financial budget, says Wong. She might skip dining at restaurants, for example, if she wants to see a friend the following week. That way, she says “people can save up for the things that matter to them.”

“The way we've posed the pandemic is that any venture outside your home is very risky, and anything inside is fine. And that, of course, is not exactly how life works,” says Monica Gandhi , an infectious disease doctor at the University of California, San Francisco, who is not involved in the project. Gandhi uses similar risk calculators in her work to assess patients’ risk of contracting various illnesses and diseases. She appreciates the site’s user-friendly interface and explanations, so people can understand the factors that sway their risk estimates.

Attaching a number to a risk can be helpful when making decisions that come with uncertainly and fear. “It tells you okay, how do you directly compare eating at a restaurant to sitting on a plane for three hours?” she says. To update the calculator and risk tracker to account for the Delta variant, the team incorporated the latest data on transmission rates and case counts. When the model was updated to include the more transmissible Delta variant, Wong noticed that activities deemed low risk a few months ago are now moderate or high risk. “I went from, ‘I can do this once a week,’ to maybe a once-every-two-months activity.”

Being within a certain range on the risk budget might seem like guaranteed safety, but that doesn’t mean your risk of infection is zero, warns F. Perry Wilson , an epidemiologist at Yale University, who is not involved in the project. Risk budgeting tools like this one can also give “a false sense of security,” he says, and “might give people sort of permission to behave in ways that might put them in more danger than they appreciate.”

Wilson is concerned about potentially wide error margins and says he is hesitant to put too much stock in a single number. Still, Wilson suggests using the tool to get a rough idea about your risk, while remembering that the studies and data the calculator is based on are still early and change often. The site’s White Paper outlines the specific peer-reviewed studies, contract tracing data and aerosol models used in the calculator, which are updated as new research emerges.

As more transmissible variants like Delta become dominant, Gandhi says these risk analysis tools may become particularly appealing to the public. “I think it's been, in a way, the hardest time of the pandemic because there was this glimpse of normalcy,” says Gandhi, referring to spring of this year. Some activities considered safe in May were deemed too risky by August. She says customizable assessment tools like the microCOVID project can be empowering to those frustrated by one-size-fits-all guidance. “Everyone has their own risk tolerance, and that may not be the risk tolerance of someone else,” says Gandhi.

The microCOVID Project is one of many Covid-19 risk assessment tools aimed at helping the public make evidence-based decisions about their health. Mathematica’s 19 and Me calculator is a credit-score style tool that lets users input their location, age, health concerns and behaviors before churning out a score that reflects risk of infection. Another freely-available tool is MyCOVIDRisk , a questionnaire-based web app that assesses the riskiness of a given scenario.

Though the team updates the microCOVID project model as new data emerges, they admit it’s challenging to stay on top of evolving information. Wong recommends that users check the calculator frequently to get the best assessment of their infection risk because local case counts can change quickly. Other pieces of the model require regular tweaking to keep estimates as accurate as possible. The team has updated the underreporting factor, which accounts for the cases that are missed in the formal case counts, three times since the pandemic began. Currently, the calculator multiplies case counts by a factor of six, assuming there are many more infected people in the area who aren’t getting tested—something Wilson thinks could have been accurate early in the pandemic, but is likely an overestimation now. Though there is a “fuzziness” to the final estimation, says Wong, it provides a framework for understanding risk.

Despite some shortcomings, Wilson thinks the calculator is particularly useful for identifying specific changes that can reduce your risk of contracting Covid-19, like adding a well-fitting mask or going outdoors. “What I found fun and informative about this was seeing how my risk rises and falls under various scenarios,” says Wilson. “Those little cues, the active things you can do [to reduce the risk of infection] I think are super useful and, and I'm glad it's out there for people to experiment with.”