Data Analysis Puts State’s COVID-19 Death Toll at 15,000 or More

COVID-19 count in NJ is skewed by several factors: testing shortages, home deaths, fatal heart attacks and strokes caused by the virus and the overwhelming task of trying to help victims
Click on a county to see COVID-19 deaths, all deaths in March, April and May 2020 and the average of deaths in those months 2015-19, as well as per capita death data. COVID-19 deaths are as of June 3, while March-May 202 data are as of June 2 and considered preliminary. Source: NJ Spotlight analysis of NJ Department of Health data.

Three months after the first COVID-19 case was confirmed in New Jersey, the official death toll stands at 11,880 — but it is likely that thousands more have died from the virus or underlying conditions exacerbated by the pandemic.

It’s impossible to say how many more, but an NJ Spotlight analysis of New Jersey Department of Health death data from 2015 through May 31 indicates that the total death toll from the novel coronavirus first diagnosed in Bergen County on March 4 could be 15,000 or more. That’s based on the dramatically higher number of deaths logged so far by state officials during the past three months, compared with the average for each month over the past five years.

“In New Jersey, in every American state, in every place in the world, the death toll is almost certainly higher than we think it is,” said Gov. Phil Murphy during Wednesday’s media briefing.

New Jersey is reporting only lab-confirmed deaths of COVID-19 for the general population and does not include probable or suspected deaths. But there is ample evidence of other deaths associated with the virus, from a shortage of testing supplies to confirm early deaths, to reports of large increases in those who died at home, to the fact that the virus attacks so many different parts of the body that medical professionals say it is at least partly to blame for some strokes and sudden deaths associated with blood clots.

Dr. Andy Berman, chief of pulmonary and critical care medicine at University Hospital in Newark, said another likely culprit is silent hypoxia — a condition that has affected some people with COVID-19 in which the blood-oxygen level is low and not enough oxygen reaches the lungs, yet a person is not having trouble breathing. This can exacerbate existing medical conditions and lead to death.

“When you’re at home, you don’t think you’re as sick as you are,” he said. “And the oxygen levels go lower and lower and that leads to some pretty serious medical consequences, especially if there’s underlying coronary disease.”

Counting COVID-19-related deaths

Christina Tan, the state epidemiologist, said Wednesday that the state is considering how to capture those other deaths in its counts.

Credit: (Rich Hundley III, The Trentonian)
State epidemiologist Dr. Christina Tan

“We are looking at a variety of different data sources to help flesh out what our death data look like,” she said. “We are also looking at the concept of probable deaths, as well.”

There is some debate among states and medical experts over the reporting of COVID-19 deaths, with currently no nationwide standard. That means the actual number of deaths, which the U.S. Centers for Disease Control and Prevention placed Wednesday at 106,202, could be higher.

The CDC changed its reporting of COVID-19 deaths on April 14 to include those deaths confirmed by a lab test and those that were likely caused by the disease because they meet clinical or epidemiological criteria, as well as those in which a death certificate lists the virus as a cause. This change was made following a position statement issued by the Council for State and Territorial Epidemiologists, according to the CDC.

About a third of the jurisdictions — states, territories and New York City — have reported more than 6,900 probable, not-confirmed COVID-19 deaths to the CDC, the agency’s website shows. The largest count by far was reported by New York City, which reported 4,755 probable deaths as of Wednesday. When the city added unconfirmed deaths to its total in mid-April, its toll increased by more than 3,700.

COVID-19 deaths, a political football

And as with everything in the current political climate, the question of whether to include probable deaths in the total COVID-19 count has become a political debate in some states. For instance, in Colorado, where the governor is a Democrat and unconfirmed deaths are included in the total, a Republican lawmaker charged that state health officials are inflating the number of viral deaths.

Politics are not behind New Jersey’s decision to count, at least for now, only confirmed deaths in its tally, officials said.

“One of the things we want to do … is to be as buttoned-down as we possibly can be, so we know, guaranteed, this is a lab-confirmed fatality,” Murphy said. “We don’t want to speculate.”

Murphy noted that the state has changed during the past months the number of long-term-care deaths it has reported. At one point, it was providing both lab-confirmed and probable deaths in nursing homes and similar facilities, but reporting the total number of deaths in all settings as only those that were lab-confirmed, which led to Murphy and others stating that more than half of the state’s deaths were occurring in long-term-care settings without being able to know that for certain.

There are roughly 1,000 additional deaths that are likely attributable to COVID-19 from long-term care facilities. The state reported a total of 6,020 resident deaths in nursing homes and similar facilities, with 4,966 of those lab-confirmed.

Testing shortage skews numbers

One reason why some deaths are not being officially attributed to the virus is that, early on, the state had so few COVID-19 tests available that it was using them sparingly. For instance, the state medical examiner released guidance on March 31 that said when one long-term care resident tested positive for the virus and another had symptoms, the second resident need not be tested. If that second resident died, the cause of death should be listed as “Probable COVID-19 Infection.”

There is evidence in the data on numbers of deaths the DOH reports daily that there could be 3,000 or more deaths in some way related to the virus.

Credit: (AP Photo/Seth Wenig)
March 25, 2020: A resident is moved from St. Joseph’s Senior Home in Woodbridge, where three patients died during an an outbreak of COVID-19.

The average number of New Jerseyans who died in March, April and May of 2015 through 2019 is 18,754. This year, the DOH is so far reporting a total of 33,770 deaths over the past three months. Between March 10, when the first death was reported, and May 31, the state reported 11,698 lab-confirmed COVID-19 deaths. That leaves 3,318 more deaths occurring in those three months than on average that are unexplained. Some 1,000 of those are most likely the unconfirmed long-term care deaths. It is not known how many of the other roughly 2,300 deaths may also be attributable to the virus — for instance, cardiac arrest that may have been exacerbated by COVID-19 — or unrelated to it.

Murphy himself noted during his June 1 media briefing that the wife of a Warren couple who died last month, Linda Auman, had tested positive for the virus while her husband Jim had not, but the family suspects COVID-19 at least contributed to his death six days after that of his wife.

“Linda’s sister said it was, I think on the death certificate, cardiac arrest but … there was lots of speculation by her sister that he (Jim) himself also had COVID positive, although he was not tested and that may well have contributed to his passing,” Murphy said.

Avoiding the emergency room

University Hospital’s Berman said that COVID-19 was likely responsible for at least some of the deaths of those who died at home, either because they didn’t realize how sick they were, couldn’t afford care or were afraid to go to a hospital to seek care because they were afraid of getting the virus there.

“Before telemedicine was really ramped up, I think people probably didn’t know how to access the medical system, other than through the emergency room,” he said.

Dr. Shereef Elnahal, president of University Hospital and former state health commissioner, said home deaths and others in which a person was not tested for the virus may not have been attributed to COVID-19 unless there was an autopsy. The increased number of deaths meant that often didn’t happen.

“It wasn’t that it was difficult; it was impossible,” he said. “Our morgue and our pathology department and those of other hospitals as well were completely overwhelmed.”

Further complicating the attribution of deaths to COVID-19 is that it is undetected in some 35% of individuals who have it, Elnahal added.

“One thing important to understand is that this disease does not manifest itself necessarily the same in everybody,” he said, noting that some patients have developed significant blood clots and some of these have led to deaths. “Some of those who have died may not have had the death attributed to the coronavirus based on how they were presenting.”

The state reported 267 lab-confirmed COVID-19 deaths in March, 6,961 in April and 4,470 in May. The full totals may not be known until state health officials look more deeply into the data, which is likely to happen as part of a review Murphy has promised will occur once officials have time to do it.

“As we do our after-action review, it will be important to get the full level of deaths,” Elnahal said.

“It takes a bit for us to look at the data,” according to Tan, the state epidemiologist. “We want to get it right before we actually share it.”