Skip to main content
ABC News
Experts Say The Coronavirus Outlook Has Worsened, But The Trajectory Is Still Unclear

The extent of the COVID-19 crisis is no more certain this week than it was last.

This is our second week following a weekly survey of infectious disease researchers from institutions around the United States. The survey asks each expert questions about the trajectory of the COVID-19 virus and they respond with three different scenarios: the most likely, the best case and the worst case. The responses are combined to produce a probabilistic consensus forecast, which can answer questions such as how many cases and deaths we should expect as well as how likely each scenario is.

The latest survey, taken on March 23 and 24, shows that experts have not become more certain or increasingly optimistic over the last week. According to the most recent survey, experts think 9 percent of all COVID-19 cases in the U.S. were known at the time of the survey. As a group, they think hospitalizations won’t peak for several more weeks or months, and they expect around 246,000 deaths related to COVID-19 in 2020.

“The situation this week is no more clear due to the diversity of responses across states, and uncertainty about the impact of those responses,” said Justin Lessler, one of the respondents and an epidemiologist at Johns Hopkins University.

But, he added, if we “adhere to guidance around social distancing, take all symptoms seriously, and self-isolate at the first sign of potential infection,” it’s possible the outcome will fall closer to the best-case scenario.

Here’s what else the researchers had to say this week:

What is the smallest, most likely and largest number of total cases that The COVID Tracking Project will report on March 29?

There is a huge amount of variation and uncertainty in the experts’ projection for the number of cases that will be reported this Sunday. The number depends not only on how quickly the virus spreads, but also on how much our testing capacity increases.

The consensus forecast suggests there will be around 117,000 cases reported this Sunday, with a 90 percent chance of having between 55,000 and 246,000.

This is up quite a bit from last week, when the consensus forecast said there would only be around 19,000 cases on March 29, though there was a change in what experts are being asked to forecast. In past surveys, the experts have been asked to forecast numbers published by the Centers for Disease Control and Prevention, which tend to be lower than those from other sources. (Also, the CDC data updates only on weekdays, so Sunday’s numbers are the same as Friday’s.) This week, the survey asks experts to forecast numbers published by The COVID Tracking Project, which aggregates data collected from state or other local public health authorities.

How many people will die in the U.S. due to COVID-19 this year?

Experts also projected a wide range of outcomes when it comes to the number of deaths that will be attributed to COVID-19 in 2020.

“The uncertainty in the number of deaths is mainly due to two things,” said Andrew Lover, a survey respondent and epidemiologist at the University of Massachusetts Amherst. “1) we’re not sure how interventions will be implemented — there are various types of stay-at-home, quarantine, lockdown interventions — and 2) if there’s a large influx into hospitals and health care workers start to get infected and can not work, things may spiral quickly.”

Lessler said that the uncertainty stems from the success of interventions, as well as how far the virus spread before it was noticed and states went on lockdown.

The expert consensus is that COVID-19 will cause 246,000 deaths in 2020, higher than last week’s estimate of 200,000 deaths. The consensus estimate ranges, though, between 36,000 and 1.1 million. That’s quite a span, with major implications for American society tucked into that uncertainty.

How do we stay at the low end of the range? “We need to protect health care system capacity by following stay-at-home and social isolation guidance — that’s really all we have at this very moment,” Lover said.

How many total COVID-19 infections were actually in the U.S. on March 23?

When the survey was conducted on March 23, The COVID Tracking Project listed about 33,000 COVID-19 cases in the U.S. Every expert surveyed thought this number understated the actual number of infections in the U.S., with one suggesting there may have already been more than 6 million infections at that time.

The expert consensus is that the real number of infections was somewhere between 83,000 and 1.8 million, with 362,000 being the most likely number.

In other words, they thought that only 9 percent of all infections had been reported, down a bit from 12 percent last week.

Of the next six months, which month will see the most COVID-19 related hospitalizations in the U.S.?

Hospitalizations in some areas of the country are rising sharply, but experts think the peak is still likely weeks or months away. When looking at the next six months, experts said hospitalizations are most likely to peak in May, according to the survey. There’s a 69 percent chance the peak will fall somewhere between May and August, experts think.

How many states and territories will report more than 100 cases by March 29?

As of Monday, there were 32 states and territories1 reporting at least 100 cases of COVID-19. By Sunday, the experts think the virus will have spread rapidly, and between 37 and 50 states and territories will have at least 100 cases, with 45 being the experts’ consensus forecast.

According to this week’s expert survey, the most likely outlook is a bit more dire than it was last week. But there’s still a wide range of trajectories, and with the proper responses, we can stay closer to the best-case end of that range.

CLARIFICATION (March 25, 2020, 5:30 p.m.): This article has been updated to more precisely reflect the language that survey administrators used in their questions to experts. The word “cases” has been replaced with “infections” in several places. In the survey, infections was used as the more inclusive term, including people who could be asymptomatic, for example.


  1. The survey included the Virgin Islands, Puerto Rico and Washington, D.C., for a maximum of 53 states and territories.

Jay Boice was a computational journalist for FiveThirtyEight.