Adding “probable” COVID-19 cases and deaths to daily totals of the ongoing pandemic has drawn the ire of critics, including President Donald Trump.
He criticized New York last week after the state began adding probable deaths — that is, the deaths of people never tested for COVID-19 but who had symptoms and possible exposure to someone who did have the disease — to its total, which in one day added nearly 3,800 deaths to the already staggering total in New York City.
But public health officials and epidemiology and infectious disease experts say using probable cases is a long-established scientific way to fully understand the impact of a disease outbreak, better respond to the disease’s impacts, and is an acknowledgement that not everyone who is infected in an outbreak will be tested.
Using probable cases and deaths “is not unusual,” Pennsylvania Department of Health Secretary Rachel Levine said during her online press conference Tuesday. “This is the way that infectious disease deaths are usually reported. But we don’t usually spend as much time delving into it” as we are now with COVID-19 being such a hot topic.
The federal Centers for Disease Control and Prevention, for example, uses probable flu cases every year to assess how a flu season is going, said Dr. Amesh Adalja, a Pittsburgh-based infectious disease physician and senior scholar at the Johns Hopkins Center for Health Security.
“During influenza season, we don’t test every person, we just start counting every person with symptoms,” he said.
Allegheny County began including probable COVID-19 cases and deaths in its daily counts this past Saturday, April 18, two days after the state began including them in its daily updates.
Using probable cases and deaths “gives us a better idea of how many people here are exposed,” Dr. Debra Bogen, director of the Allegheny County Health Department, said at the county’s online press conference Wednesday.
The county reported Wednesday that it has had 1,088 total positive COVID-19 cases so far, 36 of them probable cases, and 74 deaths, with 15 of them probable. The state reported Wednesday it has had 35,045 total positive cases, with 639 of them probable, and 1,325 deaths, with 297 of them probable.
That data will help federal, state and local governments better respond to the pandemic, Dr. Adalja said.
Using only confirmed cases of COVID-19 to do that would not make sense, he said, when “we’re responding to an outbreak in real time and deciding where resources need to go” to respond to the disease.
“It’s not going to help now to count those probable cases later,” he said.
The decision to start collecting probable cases and deaths began April 5, when the Council for State and Territorial Epidemiologists — which represents the country’s state public health officials — released a standardized definition of what constitutes probable cases and deaths.
The CDC, which often works with the CSTE to do background work like this, adopted the definition on April 14. The CDC began including probable cases and deaths in its daily updates and told state health departments to begin collecting the data.
Pennsylvania’s Department of Health then began requiring the state’s 10 county and municipal public health departments to collect data on probable cases and deaths and report that to the state. For the majority of counties across the state that do not have a local public health department, the state is investigating probable cases and deaths.
While the definition of what constitutes a probable case and a probable death are similar — symptoms and possible contact or exposure with someone who had COVID-19 — counting probable deaths is inherently less difficult than counting probable cases.
Probable deaths are identified by a “medical certifier” and listed COVID-19 as the cause of death or a contributing cause of death on the death certificate information form, said Nate Wardle, a spokesman for the state Department of Health.
But identifying a probable case takes more work, work done by members of either the state or local health departments’ contact tracing teams.
Members of the teams interview people with confirmed positive COVID-19 test results and then try to figure out everyone they may have had contact with going back to 48 hours before they got sick, said Dr. Bogen.
Often, for example, the family members who live with those people with positive cases are candidates to be counted as probable cases, said Dr. Catherine Haggerty, an associate professor of epidemiology at the University of Pittsburgh.
Being able to classify people as probable cases “is a really important part of enhancing contact tracing in real time,” she said.
Dr. Bogen said the county’s current case count includes 15 people who were identified by the county’s contact tracing team as probable cases because they were sick at home with someone who had already tested positive.
“But I suspect that number will increase a lot as we work with other teams and go back and find out about other household contacts,” Dr. Bogen said.
By counting probable cases and deaths, and not just confirmed ones, “the overall sensitivity of the surveillance system is enhanced,” Dr. Haggerty said. “Basically by casting a wider net you catch more cases. And it gives us a greater understanding of the full impact” of the disease on the community.
Though she normally researches health disparities among reproductive-aged women and their children, after the pandemic largely shut Pitt down, Dr. Haggerty was one of many epidemiologists and medical students who were asked in mid-March to help do contact tracing in Allegheny County.
She has spent entire days calling lists of people who might have had contact with a person with a positive COVID-19 case.
After she makes an assessment of each possible contact, she said she gives the information she collects to the county, which uses an algorithm to help determine if a contact should be considered a probable case.
It can be an exhausting job, she said, “but it’s very rewarding.”
Sean D. Hamill: shamill@post-gazette.com or 412-263-2579 or Twitter: @SeanDHamill
First Published: April 22, 2020, 9:51 p.m.
Updated: April 23, 2020, 3:42 p.m.
