Medical personnel prepare to test hundreds of people for the coronavirus. They lined up in vehicles last week in a Phoenix neighborhood.

Matt York/AP

hide caption

toggle caption

Matt York/AP

Medical personnel prepare to test hundreds of people for the coronavirus. They lined up in vehicles last week in a Phoenix neighborhood.

Matt York/AP

Across the United States the coronavirus is once again on the march. On Wednesday alone there were nearly 50,000 new cases — a record. The case counts for each state suggest the disease is mainly spreading in a band stretching from Florida across much of the southernmost states and westward to California, with Idaho and Iowa also in trouble.

But when you use tools to drill down to more local data, the picture gets more complicated — and even more concerning. Here are five takeaways:

It may be time for a statewide lockdown in Arizona and Florida

A key measure that epidemiologists track is the number of new coronavirus cases each day as a percentage of the population. (This per capita approach is key because it flags less populous locations that may have a large amount of spread relative to their size — and potentially more cases than their health systems are set up to deal with.)

This week a team led by researchers at Harvard came up with a rating tool with four tiers: The highest “red” alert level — triggered if a location has more than 25 new cases per day per 100,000 people — means the virus is spreading to such a degree that evidence suggests the only way to get a handle on it is to revert to stay-at-home mode. At this stage even less drastic measures such as massively ramped-up testing and contact tracing probably won’t cut it.

Technically, three states currently fall in this category: Arizona, Florida and South Carolina. But use the tool to check the situation at the county level, and the virus’s reach through Arizona in particular becomes apparent. Nearly two-thirds of the state’s counties are in the red zone — compared with about a third of South Carolina’s counties.

“I’m very worried about Arizona,” says Thomas Tsai, one of the Harvard professors who put together the tool. “We knew about the outbreak in Navajo Nation. And now we’re seeing that that entire northeast corner of Arizona, and realistically, most of Arizona, is becoming red and poses a risk of the pandemic growing out of proportion.”

Florida concerns him for the same reason. “Before the focus was really on the metropolitan areas around Miami,” Tsai says. “But now we’re seeing that counties in central Florida and even some counties in the panhandle — which had relatively few cases early on in the pandemic — are seeing an escalating rise in cases.”

There’s trouble in Texas

On the one hand Texas offers a sharp contrast. For now, transmission is mainly occurring in the state’s large cities, notes Marynia Kolak, leader of a University of Chicago team that has created another tool with county-level data. “So they have the potential to contain the outbreak before it continues,” Kolak says.

But by another measure Texas is in major trouble. A team from Dartmouth has created a tool that tracks not just how many new cases local areas are seeing per day but how quickly the overall caseload is rising each week. And the top three localities are in Texas — including the Corpus Christi area, where cases rose more than 15% over the past week. (Technically the Dartmouth tool divides areas not by county, but by communities using the same set of hospitals.)

Corpus Christi also ranks in the top 10 when it comes to daily new case counts per capita. “It’s a double whammy,” says Elliott Fisher, who spearheaded the Dartmouth tool. “It’s a place with one of the highest number of new cases and the highest growth rate. So they really need to get their act together.”

Other localities in this double-whammy category that he points to include the Victoria and Austin areas in Texas, and — in Florida — Tampa, Miami, Orlando and Jacksonville.

Hot spots are cropping up in otherwise better-off states

The local-level tools also highlight flare-ups in counties that would otherwise be overlooked because they are in states where the overall spread is not as serious.

“A cluster has to be pretty severe in order for the state to emerge as a hot spot,” the University of Chicago’s Kolak says. Yet the history of the pandemic thus far has shown that isolated flare-ups can eventually widen out with catastrophic consequences.

“Sometimes those progress and grow into larger clusters. Sometimes they don’t. So we consider those areas potential emerging risks.”

Harvard’s Tsai points to Yakima County in Washington state as one that concerns him. Other examples include counties in California and Iowa. But it’s worth noting that the vast majority of counties in the Harvard group’s red zone are not in the three red zone states.

State borders pose a challenge

The Chicago group’s tool also includes a feature that surfaces emerging regional clusters by identifying counties where the daily per capita new case number is high both there and in neighboring counties.

“What this shows is not just hot spots at a state level, but you start to notice that there’s a lot of hot spots occurring along the borders of states,” Kolak says.

Examples include between Southern California and Arizona as well as along the Mississippi River — with cases high in both the Memphis, Tenn., area and across the border in Arkansas.

This border issue points up a challenge for states, Kolak says. “There’s definite concern there, because unless you have a federally coordinated response, anything that one of those states does may not necessarily have a full impact unless the nearby states also agree.”

The choices that citizens make also play into this, she says. “So for example, if you’re in a state that has very strict social distancing guidelines, but the state next to you has very relaxed ones — driving [there] to get a haircut” could foment cross-border spread.

Ordinary citizens should check the county-level tools

While the tools described are mainly helpful to scientists and policymakers trying to detect and respond to broader trends, all of us may benefit from checking them more or less the way we look at the weather report.

Dartmouth’s Fisher says looking up the new-daily-cases-per-capita measure is a proxy for checking, “What is the likelihood that the person standing in line with you at a Starbucks [in your county] might give you COVID-19.”

Chicago’s Kolak agrees. The July Fourth celebrations are largely canceled in her area, she says. But she says she’s worried by talk that many people are planning to travel to fireworks displays farther afield.

Based on what her county-level mapping tool shows, “I’ve been encouraging friends not to go to Iowa for the July [Fourth] fireworks,” she says. “In some of those areas, cases are still pretty low. But there are counties that are starting to persistently get flagged as high relative to the nearby areas.”