Now that Gov. Larry Hogan ordered Marylanders to stay home to stem the spread of the new coronavirus, scientists are watching and waiting to see whether it works.
People’s behavior has a big effect on the computer models that show how this pandemic may unfold — and when doors to school, businesses and homes might begin to reopen and everyone may take steps back toward normalcy.
For now, there are only “reasonable best guesses at the worst-, mid- and best-case scenarios,” said Justin Lessler, an infectious disease epidemiologist at Johns Hopkins University, who’s one of many researchers working on what’s known as "planning scenarios” to guide public officials’ decision-making.
For their models, these scientists rely on such hard data as positive and negative test results and death rates in this country and others. They also consider things such things as how well people follow social-distancing orders in other countries and scientific studies that detail the properties of the virus. The models also incorporate studies and surveys that show behavior here.
The outcomes the models produce change as information changes.
There are drawbacks such as ongoing testing problems and the “massive uncertainty” from the newness of the virus, Lessler said. And there are big variables, such as human behavior.
“I think the measures will work in the short term if we do them like China; almost every model agrees on that,” Lessler said about adherence to requirements that people to stay home and away from one another.
It should be clear in a few weeks, after current infections progress and hospital cases surge and then — hopefully — peak.
That meshes with other estimates of when cases might peak in Maryland, including one from the Institute for Health Metrics and Evaluation at the University of Washington. It said a peak in hospital use could come in late April. That would be just after a projected national peak in mid-April.
The Washington model projected Maryland would have 1,766 deaths by Aug. 4, but said it could range from 338 to 4,447. It projected more than 93,500 deaths nationwide by the same date, with a possible range from 37,500 to 126,000.
At peak use of hospital beds, that model predicts Maryland will need 5,478 beds, which could explain Hogan’s focus on making 6,000 available.
Such forecasts were created to help direct resources across the country. But Hogan has been clear that there are multiple ways the outbreak could play out in Maryland.
When a state health official recently suggested a July 4 peak, Hogan immediately rebutted, saying there are various models showing different lengths of the pandemic and no one knew which offered the best prediction.
“Nobody has the exact time,” Hogan said. “There are numerous models that show four weeks, six weeks, eight weeks or longer. The federal government doesn’t have an answer. We don’t have an answer.”
A peak also doesn’t mean the pandemic is over or state directives can be lifted.
Cases would decline after a peak and there could be a lull in cases for a couple of months, but public health officials said a vaccine or drug to prevent new infections is needed before lifting restrictions. Those are in the works, but potentially some time off, especially a vaccine, estimated at a year to 18 months away.
Current problems with testing — the limited availability and lack of lab capacity and supplies — would need to be resolved before social distancing restrictions could be eased, health officials said.
Testing would need to be easily accessible for anyone, so those who are newly infected can be quickly and strictly isolated. There would need to be teams of public health workers to track down everyone a newly infected person has been in contact with.
Tests also need to be developed to identify those who have had the disease and now have the antibodies offering them some immunity. Those tests are in the works.
“It requires far more capacity for testing, both for acute ... testing and serology testing for antibodies,” said Dr. Leana Wen, a former Baltimore health commissioner and a professor at George Washington University. “Of course we need far more public health capacity to do contact tracing.”
Once testing is in place, she said, “one could imagine a scenario also where individuals who have the antibodies are issued a certificate that allows them to go back to work.”
But restrictions must remain in place until it’s clear there is a plan, said Dr. Josh Sharfstein, a former state health secretary and vice dean for public health practice and community engagement at the Johns Hopkins Bloomberg School of Public Health.
“Restrictions can only be eased when we have confidence the epidemic will not immediately surge out of control again,” he said. “What will give us confidence is the ability to identify cases quickly through testing and then reduce their chance of passing the virus to others.”
The state is working on it. It’s increasing the number of places where samples can be taken from the public for testing, including at several new drive-thru locations at state transportation hubs and hospitals.
State, private and university labs are running more than 1,000 tests a day, and the number is rising. For now, the tests require doctors’ orders and are largely limited to health care workers, seniors and other vulnerable people with symptoms, and those who are seriously sick.
State health officials say the disease modelers continue to collect data from their own data sources, as well as data from other state agencies and health care facilities in the state.
At the hospitals, the officials look at both confirmed COVID-19 patients and those who come in with symptoms.
They are counted in a system put in place in hospital emergency rooms and urgent care centers in Maryland and nationwide after the terrorist attacks of Sept. 11, 2001, called the Electronic Surveillance System for the Early Notification of Community-based Epidemics, or ESSENCE.
Charles Gischlar, a state health department spokesman, said in a statement that the ESSENCE data has been informing decisions at the state level.
“The state uses the data to assess trends and the disease burden for a measured response,” he said.
The system is used by health officials in 21 states and the U.S. Centers for Disease Control and Prevention to spot surges of flu cases each season by tracking patient symptoms rather than actual tests. As with COVID-19, most people with influenza are not tested.
ESSENCE has identified more than 87,700 people in Maryland so far this year with flu symptoms, which are similar to COVID-19 symptoms. A notable difference is trouble breathing.
Tracking specific symptoms, in the absence of widespread testing, provides another window into possible cases, said Sheri Lewis, manager of the health protection and assurance program area at Johns Hopkins’ Applied Physics Laboratory, which developed ESSENCE.
“It helps public health officials get a finger on what’s happening in their communities,” Lewis said. “People think of the system for tracking flu, but it can be used for all kinds of things. It’s been used to track Zika, SARS, Ebola. COVID-19 would be a natural extension.”
What no system has been able to do so far, however, is capture all the people who have no symptoms or weren’t sick enough to go for care, she said.
So, modelers and other scientists press on with data they have or can get, sometimes pooling the information to get a better picture. Officials at the U.S. Department of Health and Human Services, for example, recently asked hospitals nationwide to share their testing data to help federal scientists better understand the virus’ overall trajectory.
For now, many models are focusing on how to manage surges of cases rather than forecasting when they may end.
Meagan Fitzpatrick, a researcher in the University of Maryland School of Medicine’s Center for Vaccine Development and Global Health, worked with a group from Yale University and elsewhere to develop an online tool allowing hospitals to use their own patient data to forecast how things may look going forward.
A formula uses the amount of time it takes for the disease to run its course in patients to show when hospitals might run out of beds in the intensive care unit or other units. That would help hospitals make better decisions about allocating resources.
Fitzpatrick warns there are few definitive answers in forecasting the cornavirus right now. They simply offer more information to decision-makers.
There is one area of agreement, however.
“The modeling consensus now is we can’t roll back restrictions,” she said. “Once we have robust testing and surveillance we can relax restrictions. We’ll then take small steps toward normalcy.”
The model run by the University of Washington’s Institute for Health Metrics and Evaluation makes the following projections for the coronavirus pandemic in Maryland and the United States. It’s just one of many models for the trajectory of the new coronavirus.
Projected U.S. peak: mid-April
Projected Maryland peak: late April
U.S. deaths by Aug. 4: 93,500 (range of 37,500 to 126,000)
Maryland deaths by Aug. 4: 1,766 (range of 338 to 4,447)
Peak Maryland hospital beds needed: 5,478