Comparison of flu deaths with COVID-19 deaths

By Dr. Liji Thomas, MDMay 14 2020

A new study published in the journal JAMA Internal Medicine in May 2020 reports that false comparisons between the count of COVID-19 deaths as reported by various hospitals in the US, and the estimated flu-related deaths, causes significant confusion and may impair the ability to design public health policies.

By the beginning of May 2020, about 65,000 deaths had been reported in the USA as a result of the ongoing COVID-19 pandemic – a number roughly equivalent to the estimate issued by the Centers for Disease Control and Prevention (CDC) for influenza deaths this flu season.

NEW YORK CITY, APRIL 2020. As the COVID-19 coronavirus cases increase exceeding hospital capacity, mobile field hospitals such as this one in Central Park alleviate strains to treat patients. Image Credit: Robert Cicchetti

Are flu deaths comparable in number to COVID-19 deaths?

With frontline health workers pushed to the limit, and sometimes beyond, with a lack of ventilators and PPE in large cities like New York, and the unsparing demand on hospitals for weeks on end, the pandemic has placed a heavy burden on these resources unlike that experienced in the heaviest of flu seasons. Though the number of deaths in both conditions is apparently comparable, the panic and the suffering caused by COVID-19 does not compare with that of the flu, says the study.

Why, then, do health officials continue to compare the death rates between the two conditions directly? The answer may relate to a lack of understanding of how the two epidemics are reported publicly. In the case of influenza, the reported case estimates are not actual counts of patients or death, but estimates made on the basis of the specific disease codes submitted to the healthcare authorities.

Flu Death Estimates Vs. COVID-19 Death Counts

The yearly estimates of flu deaths ranged between 23,000 and 61,000 from 2013-14 to the 2018-19 seasons. However, the actual reported counts were between 3,500 and 15,600 deaths, which is almost six times less.

On the other hand, the COVID-19 mortality data is a raw count. The right comparison would be, therefore, between weekly deaths from COVID-19 and those from the flu.

For the week ending April 15, 2020, the reported number of COVID-19 deaths was 15,500, and the previous week it was 14,500. On the other hand, the CDC stats show death counts during the peak flu week in all the flu seasons from 2013-2014 to 2019-2020 stayed in the range 350 to 1,600, for an average of 752 deaths each year during the peak week of influenza.

This shows the count of COVID-19 deaths during the week ending April 21 to be anywhere from almost ten-fold to 44 times higher than the highest weekly flu death count, depending on which season is in question. It is more than 20 times higher than the average flu death count.

Even the CDC’S provisional COVID-19 death count for this week shows it to be 15 times higher than the flu deaths for this season’s peak week (ending February 29) – though the reported number is less than that obtained from other sources of data.

The Reason for The Difference

These figures help understand why frontline conditions are so appalling with the current pandemic compared to current flu death estimates. Either the latter consistently over-represents the actual number of flu deaths – or else, the currently reported COVID-19 death count is much less than the actual figure.

Which of these scenarios is more likely? The researchers contend that either is probable. COVID-19 testing restrictions and false negatives, especially in advanced illness, may lead to a falsely low number of COVID-19 deaths being reported.

On the other hand, flu death reports are notoriously unreliable because hospitals do not mandatorily have to report flu deaths in adults. The flu death figures, therefore, depend on surveillance, which are adjusted for potential under-reporting.

Are All COVID-19 Death Numbers Reliable?

Some COVID-19 death counts are becoming less reliable, as in New York City, where probable deaths from this condition are being included with the confirmed deaths, pushing the line between actual reporting and estimation of such deaths.

Some deaths included in this class may not have been due to COVID-19 at all, as when a person who has tested positive but is not seriously sick suffers a cardiac arrest. Thus, the future reappraisal of this epidemic will have to include a recount of deaths, both direct and indirect, to arrive at the number of deaths that occurred in excess of the expected. This will still be more reliable if it includes the deaths that occurred because patients with other medical conditions could not get the care they needed or if the care was delayed, because of the COVID-19-induced strain on the hospital system in severely affected areas.

Another confusing parameter is the case fatality rate. While some areas estimate the case fatality to be less than 1%, others peg it at 15%, mainly due to the limitations in the calculation. For instance, the lack of adequate test facilities means the ratio of infected persons who die to the total number of infections is falsely high since only severely symptomatic people are tested. Another source of error is the inability to include critically ill patients who were still alive at the time of counting, which falsely reduces the numerator. Serologic testing will help provide a more accurate infection number.

The most accurate and complete data set at present comes from the Diamond Princess cruise ship, which had a case fatality rate of 1.8%, reflecting 13 deaths out of 712 cases. If adjusted for the age composition, this would have been 0.5%, which is five times that of the typical case fatality rate for seasonal influenza in adults.

Implications for Public Health

In essence, direct comparisons between mortality statistics for two different conditions when the numbers are counted in two different ways lead to wrong information being relayed to the public. Secondly, when these false comparisons inform the attitude of government officials and others in power, it hinders the shaping of proper public health policies. For instance, such comparisons may increase the tendency to accelerate the reopening of the economy while relaxing movement and social restrictions.

The researchers summarize their conclusion, peremptorily, “Although officials may say that SARS-CoV-2 is “just another flu,” this is not true. Our analysis suggests that comparisons between SARS-CoV-2 mortality and seasonal influenza mortality must be made using an apples-to-apples comparison, not an apples-to-oranges comparison.” This will allow the right public health interventions to be made to mitigate the health and economic impact of COVID-19.