Elizabeth Rosenthal wrote an insightful and emotional article about the death of her mother from COVID-19. She explains that her mother’s death certificate stated “Acute Respiratory Distress Syndrome” as the primary cause, and the secondary was “suspected COVID-19.” The White House, the Centers for Disease Control and Prevention and the states are debating the proper theoretical (and politically beneficial) way to tally COVID-19 deaths.
“Though my mother almost certainly died of COVID-19 (she met the clinical case definition), her death was, as far as I can tell, not counted — and certainly will not be counted if the White House gets its way. Unfortunately, counting COVID deaths and cases has been turned into a battle of semantics, chance, bureaucracy, politics and immediate circumstance, rather than science.”
First of all, states, agencies and workplaces are all counting differently, sometimes bending, I suspect, to political convenience: At first both the CDC and New York state required a positive test to count as a COVID death. Later, each decided to add, for some purposes, cases from senior care facilities in which “presumed” or “assumed” COVID-19 was listed on a death certificate, or when a doctor deemed it was the probable cause of death.
But what of “suspected COVID-19,” the term on my mom’s death certificate? Almost certainly not. Jill Montag, a spokesperson for the New York Department of Health, explained: “The case count for deaths of residents in nursing homes and adult care facilities includes presumed (as determined by a physician) COVID-19 deaths.”
And how did the hospice doctor, who had never physically seen my mother, determine what word to use? Did she mean to draw a distinction between “presumed” and “suspected”? As a doctor who filled out these forms during the AIDS epidemic in New York, I know it’s often a quick and arbitrary process, not bound by rules; generally, the primary cause is listed as cardiopulmonary or respiratory arrest, with secondary and tertiary causes after that. If a patient also had pneumonia and diabetes, COVID-19 might not even make the cut.
Indeed, on May 16, Colorado said it would not count nearly 300 people who had tested positive but did not have COVID-19 listed on a death certificate.
Let me explain how the coronavirus crisis played out in my mother’s final month of life: Shortly after she turned 96 in March, she spiked a temperature of 102, raising alarms. Although the facility had gone into lockdown weeks earlier, we knew it had several confirmed cases among residents. The doctor first checked her urine and skin. All were clear. But a chest X-ray showed bilateral pneumonia, and the next day she had a dry cough. Exposure+fever+cough+bilateral infiltrates = COVID-19.
Had she gone to a hospital and needed admission, she would have been tested for COVID-19 and counted. But she didn’t want aggressive treatment, such as a ventilator, and she was not in pain or struggling to breathe. So she stayed in her apartment.
Because my mother was “presumed COVID,” the geriatrician started standard outpatient COVID treatment: azithromycin and supplemental oxygen, as needed.
But the third reason she was not tested is, from a public health standpoint, disturbing: Many of the people who might have mandated or performed a COVID test at that point do not really want to know if patients like her have the disease.
Not testing or transparently reporting COVID cases is a great way to keep numbers low. As Trump said when he balked at having American passengers come ashore from an infected cruise ship on March 7, “I like the numbers being where they are.”
For states, avoiding robust testing and reporting is a good way to make sure new cases decrease for 14 days (a CDC recommendation for reopening). But it deprives those same states of crucial information for rational decisions.
The data scientist who designed Florida’s COVID reporting system was fired recently because, she said, she had made it too transparent for her superiors in the state’s health department.
By not testing residents like my mom, her facility could report for weeks that it had only four positive cases on the premises — even though that might not reflect the underlying reality.
What we need instead is a single and clear national testing strategy outlining who must be tested, when and with which test.
My mom was 96, frail and declining, though she had no chronic conditions per se. COVID-19 was not a terrible way for her to pass. But that doesn’t mean her death shouldn’t be counted for what it was. Cataloging deaths like hers would make our “numbers” look worse, yes. But it would also greatly add to our knowledge of how this virus spreads and affects patients, as we try to develop a coherent strategic reaction.
My mother had a long and meaningful life. Being counted would give her premature death meaning as well.