Publicly reported deaths caused by the coronavirus (COVID-19) may be misrepresenting the death toll of the pandemic, according to research recently published in JAMA and recommended by our Faculty of experts.

The author Steven H.Woolf and co-authors estimated the excess deaths during the early weeks of the pandemic and the relative contribution of COVID-19 compared to other causes. Today we had the chance to speak to the author, Steven, for further insight into this important research.

Can you explain why you and your co-authors felt it was important to pursue this aspect of research into COVID-19?

We run a center that studies population health and health inequities. For years we have examined evidence about the social and economic factors that affect health and about the population groups and places (down to the census tract) with greater risks for disease and premature death. When the pandemic struck and it became clear that society was about to experience a major disruption in daily life, we began to worry about the scale of the impact and we predicted that the number of deaths from the pandemic would exceed the number of deaths officially attributed to COVID-19.

Excess death calculations are an accepted method for gauging the scale of a public health emergency. It was recently used to estimate the number of deaths in Puerto Rico from Hurricane Maria. We—as well as researchers at other universities, the CDC, and news outlets—turned to this method to estimate how many total deaths the pandemic had caused.

As we look ahead to the danger of a winter resurgence in COVID-19 and see cases rising across much of the nation, our study urges that we learn from past mistakes and lean forward in enforcing masks, social distancing, and avoiding large gatherings and venues.

Can you tell us about the particular challenges you came upon when conducting your research? 

The biggest challenge is that we are forced to work with provisional data. It normally takes the CDC 1-2 years before they can release reliable mortality data or details on specific causes of death, but because of the current emergency, they have rushed out provisional data, which are inherently incomplete. Causes of death are lumped into broad categories called “grouped causes,” such as cancer, and we can’t dig into data on specific causes such as breast cancer, drug overdoses, or suicides. A particular concern for this study is the coding of COVID-19 deaths, which were listed in different sections of the death certificate. We didn’t get bogged down in the debate over the appropriateness of where the virus was listed on the death certificate but counted COVID-19 deaths as any death in which the virus was listed on the death certificate. Of interest to us was how many excess deaths occurred with no mention of COVID-19.

What are the dangers of underestimating deaths caused by COVID-19?

The main reason to identify the undercount is to make sure the public is properly informed about the scale of the impact. We live in a time in which politicians (and conspiracy theorists) are downplaying the impact of the pandemic or claiming that the death count is exaggerated. As more studies like ours accumulate, the evidence suggests quite the opposite, that the number of deaths we hear on the news is an underestimate of the total loss of life. The other reason to recognize excess deaths is to draw attention to the plight of people who were not infected by the virus but died from other causes because of disruptions caused by the pandemic. Understanding where and why that is happening is important so that the attention to coronavirus doesn’t steal attention and resources away from other people with life-threatening conditions.

What do you hope your paper will add to the efforts of combating COVID-19?

Showing the outsized impact of the pandemic will hopefully sound an alarm about the urgency of policy actions to control the pandemic, at a time when too many in our public and leadership are wanting to back off. Our updated study, published on October 12, shows that states that were hit first by the pandemic—such as New York and New Jersey—experienced a huge spike in deaths in mid-April but were able to mount a robust response and get themselves back to baseline by May; 90% of all excess deaths occurred within 8-9 weeks. In contrast, states that lifted restrictions early and reopened the economy in late April or the first week of May, such as Florida and Texas, experienced a protracted surge that extended through the summer. Our analysis cut off at the end of July, but by then their surge had already lasted 16-17 weeks. As we look ahead to the danger of a winter resurgence in COVID-19 and see cases rising across much of the nation, our study urges that we learn from past mistakes and lean forward in enforcing masks, social distancing, and avoiding large gatherings and venues. We know this issue is caught up in politics, but people who protect their “liberty” by flouting this guidance, thinking they are rescuing the economy, not only jeopardize the health of themselves and their loved ones but also prolong the devastation to the economy. We can’t save lives or save the economy without getting this virus under control.

The pandemic has shown the solidarity of the scientific community, in which colleagues across the globe leapt into action and worked together to share data and make discoveries at breathtaking speed to understand the virus, invent and test treatments and vaccines, and rush information to physicians who updated their COVID-19 protocols in real-time.

Your study suggests 25,000 excess deaths in New York City, compared to the reported 18,600 COVID-19 deaths. Which means that 6,000-7,000 deaths were not reported – can you comment on any reason for this discrepancy between the numbers?

This question is not unique to New York City; excess deaths for most of the country exceeded COVID-19 death counts. To be clear, it’s not that the deaths “were not reported.” We are actually using reported deaths to calculate “excess deaths.” We arrive at excess death by comparing the reported deaths with the number that would be expected in a normal year. The COVID-19 death count explains about two thirds (67%) of these excess deaths. The remaining third comes from two sources: people who died of COVID-19 but—for various reasons—the virus was not listed on their death certificate, and people who died of causes unrelated to the virus itself but who succumbed from circumstances brought on by disruptions of the pandemic. These could be deaths due to acute emergencies like a heart attack, chronic diseases like diabetes, or mental health or substance abuse crises that led to overdoses or suicides.

Can you comment on the limitations of your study?

As noted earlier, the use of provisional data comes with its own limitations, and death certificates are subject to errors and inconsistences across medical examiners and doctors who complete them. We rely on a statistical model to calculate excess deaths, and any model relies on assumptions. Change the assumptions and you can change the results, although we’ve used sensitivity analyses and found that the results are robust across different assumptions. It’s also possible that the excess loss of life is greater than we think because all we can see is the “net” excess deaths. The pandemic has decreased mortality from some causes, such as car accidents and possibly the flu, which may offset increases in excess deaths and dampen the “net” result we are seeing.

Excess death calculations are an accepted method for gauging the scale of a public health emergency. We—as well as researchers at other universities, the CDC, and news outlets—turned to this method to estimate how many total deaths the pandemic had caused.

How has your work and life been disrupted by the pandemic?

Like so many Americans, we are doing this research from our homes, meeting virtually to talk with colleagues, and doing so steps away from our families and barking dogs. We are far more fortunate than so many workers, who are on the front lines and putting themselves at risk, and than so many of our fellow Americans who have lost their jobs, their homes, and their financial security.

This pandemic has caused thousands of deaths and misery across the globe, but can you comment on any positives which have come out of this period?

There is a silver lining in many aspects of life, such as coming to appreciate the things we take for granted, but I will focus my response on science and policy. The pandemic has shown the solidarity of the scientific community, in which colleagues across the globe leapt into action and worked together to share data and make discoveries at breathtaking speed to understand the virus, invent and test treatments and vaccines, and rush information to physicians who updated their COVID-19 protocols in real-time.

For policy, my wishful thinking is that this catastrophe will underscore the influence of policy on health—a linkage I have spent much of my career trying to explain. The pandemic has shown in vivid terms that what politicians do—or resist doing—can change the fate of a population and, when handled poorly, can claim thousands of lives. In the United States, we have seen how outcomes across states vary depending on the choices (and politics) of a governor, allowing people in one state to live longer than others across state lines.

Our Faculty have evaluated the literature relevant to the recent COVID-19 outbreak and recommended the most relevant and important research to help tackle this global crisis. We have made these freely available here.

Hannah Towfiq

Hannah Towfiq

Hannah is the Social Media & Content Manager at Faculty Opinions and Sciwheel. Her background is based in science and healthcare, which she utilises to communicate complex subjects in insightful and approachable formats, through our blogs and social platforms.

Leave a Reply