Now that Georgia is reopening, teams of experts are calculating whether the rate at which people will sicken or die from the novel coronavirus will rise in the coming weeks, and whether we'll need more hospital beds to care for them.

The news is generally bleak. Revised projections show that unless people act as if they remain on lockdown, infections and deaths by COVID-19 are bound to rise. The only disagreement among experts is by how much and when.

Even a small bump in how much Georgians move about their communities could cause the state’s death toll to tick upwards. A new projection by the Institute for Health Metrics and Evaluation, a model favored by the White House, assumes many Georgians will stay at home and summer heat will suppress the virus. Still, IHME revised its cumulative death toll figures upwards to nearly 5,000 by August, an increase of more than 140% percent.

» COMPLETE COVERAGE: Coronavirus in Georgia

Other models show that if Georgians return closer to their pre-pandemic ways, the state could experience a devastating, exponential spike of infection similar to those in hard-hit areas such as New York City. One created through a collaboration between Massachusetts General Hospital, Harvard Medical School, Georgia Tech and Boston Medical Center projects that if restrictions end, cumulative deaths will most likely creep up through July, then skyrocket well beyond 10,000 the next month if residents stop sheltering in place after May 25.

Cell phone and other data, however, already show that Georgians have begun to travel more often.

The Georgia Department of Public Health is preparing for different scenarios by working with modelers from Georgia Tech, Massachusetts Institute of Technology, IMHE, and the University of Pennsylvania to project local health care needs, a spokeswoman said. It plans to share the results with hospitals and local health agencies.

But even top experts have not rehearsed this kind of crisis, said University of Texas at Austin Professor Lauren Ancel Meyers, who researches ways to project the spread of disease using data science and statistics.

“We did not have a play book for this,” Meyers said.

On Wednesday, DPH logged 898 new confirmed COVID-19 cases and 23 new deaths.

Sheltering worked, experts say

Projections for the spread of the novel coronavirus vary and there is great uncertainty in each of them. But the researchers who made them agree that sheltering in place worked for Georgia. The spread of the disease slowed.

What epidemiologists call the reproduction number, or “R,” sank to about 1, according to estimates from a Columbia University team; the collaboration involving Georgia Tech; and IHME.

This means that towards the end of the state’s lockdown, each person who contracted the virus spread it to about one other person — a key threshold for containing it.

» RELATED: Lifting stay-at-home order, Kemp shifts focus to economic recovery

» MORE: Georgia prepares to launch new coronavirus tracing program

If R remains even slightly above 1, the disease can overrun a health care system over the course of several months. If it remains at 1 it never goes away. If it dips below 1, the number of new cases is shrinking and the disease can be stopped.

“It was a clear indication it was working, that sheltering in place was the right decision,” said Turgay Ayer, a Georgia Tech professor and expert on healthcare analytics who was part of the Tech collaboration.

Experts in infectious disease agree that the further below 1 that R drops, the easier it is to keep a disease in check. If a state has only a handful of new cases per day, it can find the manpower and supplies to isolate each infected person and test their contacts.

This approach worked so well in South Korea that the country of more than 51 million put the novel coronavirus in check while businesses remained open, observed Columbia University Professor Jeffrey Shaman, an expert in infectious disease projections who is with that college’s team of COVID-19 modellers.

One reason South Korea’s approach is so powerful is that it identifies and isolates people who are infected before they show symptoms, keeping them from unwittingly spreading the virus.

“They’ve really throttled the disease,” he said.

A lag before a surge

Had sheltering restrictions continued in Georgia, the number of new infections would have likely gone into decline. Public health experts recommend reopening when R is well below 1 — perhaps at 0.5 or below — so that decision makers have a better shot at detecting whether relaxing restrictions is increasing the spread of the disease before it spreads out of control.

Rather than wait, Georgia began to reopen.

“Relaxing the measures at that level — that’s premature,” said Ali Mokdad, a professor at the University of Washington’s IHME. “That leads to continued circulation of the virus.”

Those working to control the spread of the disease are now in a bind, said Shaman. Test results cannot tell public health officials if cases are rising or falling in the present moment. It can take a week or more between when a person is infected, begins to infect others, shows symptoms, undergoes testing and gets the results.

"We did not have a play book for this." —Lauren Ancel Meyers, University of Texas at Austin Professor

This means that any spike in the spread of the disease will only show up in Georgia's case tracking data a week or two after it is underway, which may be too late. In the New York metro area, cases spiked from below 200 to more than 50,000 in less than 20 days.

“For two weeks you’ll have no indication that cases are rising again. You won’t know it,” Shaman said.

The model by Shaman’s team shows that Georgia infections will begin to accelerate towards the end of this month, while most states that continue restrictions such as Michigan will experience a drop.

Preventing the worst

What leaders must do now is to heed the models, experts said. None can predict the future, and there is great uncertainty in their projections. But they can help leaders prevent the worst from happening, and prepare them for the likely consequences of their policies.

As Georgia opens up, disease modelers recommend building up contact tracing and isolation efforts, which the state is doing currently.

University of Texas’ Meyers added that states should boost efforts to protect vulnerable seniors in nursing homes and elsewhere. People who must continue to work can receive paid sick days so they don’t say at work and spread the disease. Local leaders can stay in daily contact with researchers as they do in Austin, learning what measures to track and how to turn them into effective policy.

IHME’s Mokdad said the medical system must be prepared for a second wave of the novel coronavirus, which he predicts will take place during flu season. Workers at chicken processing plants and other key points in Georgia’s food production system should also be protected.

“You don’t want to add on top of COVID-19 high prices and a shortage of food. That would be a disaster,” Mokdad said.