Researchers from Columbia University’s Mailman School of Public Health created an algorithmic model to glean a more accurate picture of the virus’s spread than was conveyed by data reported by states and locales to the Centers for Disease Control (CDC) that was weakened by undercounts of active cases and incomplete information that didn’t incorporate metrics such as breakthrough COVID infections or patients who were sick but didn’t need to be hospitalized. According to their research, the true number of cases could be about five times the number of confirmed cases tracked by Johns Hopkins University last year.

“The vast majority of infections were not accounted for by the number of confirmed cases,” said Columbia University professor of environmental health sciences and one of the study’s authors, Jeffrey Shaman, of the findings. “It is these undocumented cases, which are often mild or asymptomatic infections, that allow the virus to spread quickly through the broader population.”

The Columbia team’s model simulated the coronavirus’s spread in 3,142 U.S. counties and considered factors such as how many people may move between locales in a given time. It analyzed five major metropolitan areas in the U.S. including Los Angeles, New York, Chicago, Miami, and Phoenix, as well as data from across the country, and found that three separate waves of COVID had lopsided effects in 2020 depending on the season and your physical location.

For instance, the first major set of cases stemmed from densely populated urban areas in states such as New York and California, while the summer ravaged Southern states and parts of the Midwest. Infections saw a lull in New York and Chicago in the summer months while Los Angeles experienced a new wave during the same time. In states across the upper Midwest and the Mississippi Valley, more than 60% of the population had had COVID by the end of 2020, according to the study, while major metropolitan regions saw between 44% and 52% of their residents infected.

But the more telling number going forward is the virus’s contagion rate. Researchers found that about 0.77% of the U.S. population, or one in 130 people, were actively contagious at the close of last year even if they weren’t all dying or in the hospital, and that there was an even higher share of contagious people in urban and metropolitan areas. That doesn’t even account for so-called latent infections where someone is infected with the pathogen but not yet contagious.

COVID vaccines and stricter public health measures can help make a significant dent in those figures by the end of 2021. But challenges remain owing to inconsistently enforced safety measures across the nation and strains like the Delta variant that may present more of a challenge for certain coronavirus jabs. There’s also the question of how long, and how strongly, COVID immunity persists over time as antibody levels begin to dwindle—and how much Americans will embrace booster shots as they become more widely available.

“While the landscape has changed with the availability of vaccines and the spread of new variants, it is important to recognize just how dangerous the pandemic was in its first year,” said Columbia assistant professor and study author Sen Pei. In their pandemic prognosis for 2021, the researchers predict the coronavirus will continue to spread among the previously uninfected as breakthrough and mild cases continue to hamper public health efforts by facilitating this spread.

There have been more than 38 million coronavirus cases reported in the U.S. to date, according to Johns Hopkins, and per Department of Health and Human Services (HHS) hospital utilization data, the massive surge of Delta variant–linked infections in the past few months pushed the number of people currently hospitalized with COVID past 100,000 for the first time since January.