When I look at COVID-19 data, I see very large disparities in the volume of cases and deaths between various countries as well as between states within the U.S.

Many factors, including time of the testing ratio, time since outbreak, weather, population density, age of the population, health of the population, government decisions, and personal choices can impact the virus’s severity in a region.

Another reason why it’s difficult to compare regional severity is the lack of a universal definition of what constitutes a COVID-19 case or cause of death.

The same holds true within America. The Centers for Disease Control and Prevention (CDC) has no clear singular definition of what is or is not a COVID-19 death. Death certificate guidelines published by the CDC on March 4 state:

It is important to emphasize that Coronavirus Disease 2019 or COVID-19 should be reported on the death certificate for all decedents where the disease caused or is assumed to have caused or contributed to death.

The “assumed to have caused” and “contributed” instructions are not standard practices for other diseases. The result is states, and even care facilities, have chosen different methods of documenting COVID-19 deaths. This cannot help but inflate COVID-19 statistics compared to other causes and increase confusion.

During Gov. JB Pritzker’s April 19 health briefing, Dr. Ngozi Ezike, the Illinois Department of Public Health (IDPH) director, gave the state’s definition of a COVID-19 death. Simply put, it means at the time of death, a person had a COVID-19 diagnosis.

“Technically, even if you died of a clear alternate cause, but you had COVID at the same time, it’s still listed as a COVID death. So, everyone who’s listed as a COVID death doesn’t mean that was the cause of death, but they had COVID at the time of death,” said Ezike.  As an example, she said, “If you were in hospice and had already been given a few weeks to live, and then you also were found to have COVID, that would be counted as a COVID death.”

In some states, like Alabama, the deceased must have tested positive for COVID-19 for the death to be reported as a COVID-19 death. Other states, like New York, allow for more leeway. A death may be reported as a COVID-19 death if the deceased was suspected of having it even if no test was performed. Some states require COVID-19 to be the primary cause of death in order to qualify for state reporting, while others report all cases where COVID-19 was present at the time of death to be COVID-19 deaths. These divergent methods make some region’s infection and death rates undercounted and others overcounted, making it difficult to gauge the disease’s true impact on medical resources and mortality rates. This is especially problematic when death counts influence hospital reimbursements and federal relief.

According to a recent Washington Post article, Alabama officials have ruled that one of every 10 people who died with COVID-19 did not die of COVID-19.

In Colorado, the number of COVID-19 deaths recently dropped by 24% The Colorado Department of Public Health and Environment (CDPHE) said it previously counted deaths of persons who had COVID-19 in its totals, even if COVID-19 was not the cause of death. In one example, the Montezuma County coroner found that a man who had died of acute alcohol poisoning – his blood alcohol level was nearly double the lethal level – had his death classified as COVID-19. In revising its reporting policy, the CDPHE will now distinguish between deaths caused by COVID-19 and deaths where COVID-19 was present. “What the people want to know is not who died with COVID-19, but who died of COVID-19,” said Gov. Jared Polis.

In states where deaths can be attributed to COVID-19 without confirmed diagnosis, those deaths can be counted in the numerator (people that died from COVID-19), but not in the denominator (people that tested positive for COVID-19) of their death ratios. This is fundamentally flawed math.

When experts look back on this time, I think not standardizing the definition of a COVID-19 death will be considered a significant error on the part of the U.S. and the world in trying to understand and control this pandemic. Overreactions or underreactions are certain given the disparity in methods. Flawed and inconsistent data undermines credibility. It is really important when digesting data to know how it is calculated and to not assume data from one source is actually comparable to another source.

Stay safe, stay strong!