Mass. coronavirus peak likely in mid-April, some epidemiologists predict

Some researchers caution that new incoming data can modify predictions

Medical workers tested for the coronavirus outside at CVS in Shrewsbury.Matthew J. Lee/Globe Staff

New forecasts for the spread of coronavirus predict that stress on Massachusetts hospitals will peak around April 14, and that by summer the deadly virus will have taken nearly 1,800 lives in the state.

These projections by the University of Washington forecast that COVID-19 will kill more than 80,000 people in the United States by mid-July. At its peak in mid-April, the virus will kill as many as 2,200 Americans a day, the model predicts.

Those numbers presume that the United States will maintain a policy of social distancing through May, said Christopher Murray, director of the University of Washington’s Institute for Health Metrics and Evaluation, which publishes the forecast and will be regularly updating it as new data comes in.

“If we don’t social distance through May . . . things will be worse,” Murray said in a telephone interview.

Echoing the Washington researchers’ prediction, Governor Charlie Baker on Monday said the state’s epidemic would likely start to peak between April 7 and 17. In preparation, he said, the state has moved to add 1,000 nursing home beds, recruit retired clinicians, cancel elective surgeries, and request the federal government send 1,000 ventilators, which it agreed to do.

“We’ve been doing a lot of things through social distancing and other mechanisms to bend the trend on that surge,” Baker said. “But the mere fact that we’re anticipating that we need 1,000 ventilators, and the feds have basically agreed to that . . . would imply that we think that it’s going to be a pretty big deal.”

Massachusetts has recorded 5,752 cases of COVID-19 and 56 deaths from the disease, as of Monday. The grim report from biostatisticians at the University of Washington predicts that 1,782 people in Massachusetts will die from the virus by this summer.

Murray is one of many experts trying to divine the path of the coronavirus pandemic, to provide doctors and policy makers some guideposts for decisions on how to combat the spread of the disease and where to disperse finite resources to treat the infected.

Virtually all of the studies conclude that the worst is yet to come, and that the peak is roughly two or three weeks away.

If the United States maintains social distancing, the model predicts that deaths nationally from coronavirus will drop to fewer than 100 a day by early June, according to the UW model.

“I think by the time the beginning of June rolls around, we will be at the tail, if the social distancing stays in place and if — the biggest if — if [social distancing] works like it did in China,” Murray said.

The University of Washington numbers are slightly more optimistic than estimates offered over the weekend by Dr. Anthony Fauci, from the White House’s coronavirus task force, who suggested the coronavirus could kill 100,000 to 200,000 Americans.

The UW modeling project started about two weeks ago, when the hospital system at the University of Washington asked for a forecast to help it plan for how many beds and ventilators the system might need. Other hospitals heard about the project by word of mouth, and they asked for help, too.

“And so then rather than go piecemeal — this hospital system, that hospital system — we just decided we’d do it to all 50 states,” Murray said. The first version went live last Thursday and was updated on Monday.

“And now because we’re just getting a flood of better data — more inputs, more information on intubation rates, et cetera — we will keep it up-to-date, and we will expand on Wednesday to Europe and then sometime by the weekend to the rest of the world.”

About 60 experts are working on the model, he said.

As more data about the effects of the virus becomes available, coronavirus models should get better. Murray said people who study COVID-19 are this week carefully watching the numbers from hard-hit Northern Italy.

“Will the regions in Italy — like Lombardy and Liguria that were ahead of us in this epidemic and which put in place social distance measures — will they peak about the time we expect them to peak?” he said. “That should be this week. So everybody is watching what happens there and that’ll give us some insight as to whether social distancing as implemented in Italy is about as effective as Wuhan [China] or much less effective. That’s a critical moment for us this week to see what happens.”

Helen E. Jenkins, professor of biostatistics at the Boston University School of Public Health, has not studied the University of Washington model closely, but explained that disease models, in general, are based on data such as the number of cases of the virus in a region, the average number of people to whom an infected person might pass the virus, and the number of people who may have contracted it. It also looks at how many people have recovered and may now have some resistance to the disease, among other factors.

“It gets difficult pretty quick,” she said. “Obviously I am a big fan of modeling, but I think that sometimes, especially in the media, an awful lot of emphasis is placed on a number that comes out through a model without necessarily understanding that it's very dependent upon the data,” which can be incomplete or imprecise in the midst of a worldwide pandemic.

“We shouldn’t put too much weight on any single number,” she said.

To be sure, specialists acknowledge they aren’t sure about many aspects of the virus’s trajectory. Confirmed cases are helpful, but not always a true indicator of an outbreak’s size, as many sick people may go untested. And it’s unclear how long people who recover from infection are immune.

Even so, the new model’s findings are consistent with evidence from South Korea that showed it took about two to three weeks after social distancing measures for confirmed cases to fall, said Samuel Scarpino, a Northeastern University assistant professor who specializes in infectious disease modeling. Hospitalizations tend to fall one to three weeks later, he said, due to the illness’s progression.

But it’s unclear when that clock started in Massachusetts, as restrictions came in waves. Since the first measures were imposed in Boston around mid-March, location data shows that movement among people fell by 50 percent, a bigger drop than occurs on many holidays, Scarpino said.

“We still don’t know the answer to exactly what kind of physical distancing measures have to go into place before we start counting down,” Scarpino said. “Is it the full-scale lockdown that we saw in Italy or is it the measures in South Korea, which look a little more like what we’re doing in Boston?”

The only way to know when the peak has hit is when illnesses fall for one to two weeks straight, Scarpino said. Before leaders consider relaxing restrictions, he said, they should learn from what worked in South Korea, Singapore, Hong Kong, and Taiwan — and ensure that testing remains widespread and that infected people are isolated, and their contacts possibly quarantined as well.

That’s a key part of a new plan by public health experts at the American Enterprise Institute led by Dr. Scott Gottlieb, the former US Food and Drug Administration commissioner and an adviser for Baker on the virus.

The AEI plan says states should only relax restrictions when they reach three milestones: confirmed cases fall for at least 14 days, which is how long the virus can take to cause symptoms; local hospitals are able to safely treat all coronavirus patients; and the state has the capacity to test anyone with coronavirus symptoms.

Afterwards, schools and businesses can reopen, the report says, but gatherings should remain limited and people over 60 or with underlying medical conditions should “limit their time in the community.”

Heightened testing and vigilance will be necessary to prevent new outbreaks from exploding into larger epidemics until a vaccine is available in a year or more, said Alex Perkins, a University of Notre Dame epidemiologist.

“The trick is figuring out how we make some kind of adjustments going forward that minimize the risk but still enable economic activity and other aspects of normal life to resume,” Perkins said.

For now, epidemiologists warn people to keep following the new rules which should start to pay off. “People should be prepared,” Scarpino said. “We’re still headed toward the worst part of this.”