The Limitations of Surveillance - Credit CDC
#15,604
Trying to determine the number of deaths, and the CFR (case fatality rate) during a pandemic is a daunting task, partially because these numbers can fluctuate over time and location, and because preliminary data is often inaccurate or incomplete.
Scholars are still arguing over the actual death toll in 1918 (20 to 80 million), and the 1957 pandemic has been described as killing anywhere from 1 to 4 million people, although the CDC cites: The estimated number of deaths was 1.1 million worldwide and 116,000 in the United States.
We are already well beyond the 1957 numbers, and destined to go much higher. While we are at just over 280,000 counted deaths in the United States, excess mortality since the arrival of the virus suggests that number could be closer to 400,000.
Since we can't count and verify every single flu death (or by any other infectious disease), we have to rely on a variety of statistical analyses to come up with a reasonable estimate (see Apples, Oranges, And Influenza Death Tolls).
While no one analysis can be said to be `right', when taken together, they can at least put us in the right ballpark. And the 5 to 10 times deadlier than influenza - at least in adults over 40 - is a range we keep hearing.
Adding to our knowledge of the impact of COVID-19, we have the following study out of Denmark, which estimates that the likelihood of dying from COVID-19 within 30 days of diagnosis is roughly 5 times greater than with influenza.
It's as long, detailed, and technical open-access report, so I'll simply post the abstract and urge those who are interested to follow the link.
Thirty-Day Mortality and Morbidity in COVID-19 Positive vs. COVID-19 Negative Individuals and vs. Individuals Tested for Influenza A/B: A Population-Based Study
Vardan Nersesjan1†, Moshgan Amiri1†, Hanne K. Christensen2,3, Michael E. Benros4,5*‡ and Daniel Kondziella1,3*‡
Background: As of October 2020, COVID-19 has caused 1,000,000 deaths worldwide. However, large-scale studies of COVID-19 mortality and new-onset comorbidity compared to individuals tested negative for COVID-19 and individuals tested for influenza A/B are lacking. We investigated COVID-19 30-day mortality and new-onset comorbidity compared to individuals with negative COVID-19 test results and individuals tested for influenza A/B.
Methods and findings: This population-based cohort study utilized electronic health records covering roughly half (n = 2,647,229) of Denmark's population, with nationwide linkage of microbiology test results and death records. All individuals ≥18 years tested for COVID-19 and individuals tested for influenza A/B were followed from 11/2017 to 06/2020.
Main outcome was 30-day mortality after a test for either COVID-19 or influenza. Secondary outcomes were major comorbidity diagnoses 30-days after the test for either COVID-19 or influenza A/B.
In total, 224,639 individuals were tested for COVID-19. To enhance comparability, we stratified the population for in- and outpatient status at the time of testing. Among inpatients positive for COVID-19, 356 of 1,657 (21%) died within 30 days, which was a 3.0 to 3.1-fold increased 30-day mortality rate, when compared to influenza and COVID-19-negative inpatients (all p < 0.001). For outpatients, 128 of 6,263 (2%) COVID-19-positive patients died within 30 days, which was a 5.5 to 6.9-fold increased mortality rate compared to individuals tested negative for COVID-19 or individuals tested positive or negative for influenza, respectively (all p < 0.001). Compared to hospitalized patients with influenza A/B, new-onset ischemic stroke, diabetes and nephropathy occurred more frequently in inpatients with COVID-19 (all p < 0.05).(Continue . . . .)
Conclusions: In this population-based study comparing COVID-19 positive with COVID-19 negative individuals and individuals tested for influenza, COVID-19 was associated with increased rates of major systemic and vascular comorbidity and substantially higher mortality. Results should be interpreted with caution because of differences in test strategies for COVID-19 and influenza, use of aggregated data, the limited 30-day follow-up and the possibility for changing mortality rates as the pandemic unfolds. However, the true COVID-19 mortality may even be higher than the stated 3.0 to 5.5-fold increase, owing to more extensive testing for COVID-19.
