Caveat Emptor--always be critical when reading statistics.


In a recent JAMA article on the leading causes of death for 2020, Covid19 is ranked fourth with a total of 345,323 deaths trailing third behind Cancer (598,932 deaths), and Heart disease (690,882 deaths). The source of this data was from the CDC (http://www.cdc.gov/nchs/deaths.htm). Note that unfortunately in 2020, we see that deaths related to Diabetes increased by 15% and deaths related to Alzheimer's disease increased by 9.7% both in a single year!

The number of covid deaths is based on what hospitals recorded as the underlying cause of death for patients. This means that on the death certificate the original reason you came into the hospital would be counted as the cause of death. In the US at least, it seems that the number of deaths due to Influenza and pneumonia has remained relatively stable between 2019-2020, so it gives us some confidence in the numbers reported for Covid19.

Are there secondary co-morbidities listed with Covid-19 deaths?


Almost half of Covid-19 deaths have pneumonia or the flu listed as a comorbidity and we see that hypertension, diabetes and Alzheimer's disease were the most common comorbidities when patients came in with Covid-19 symptoms.

What about a breakdown by race and ethnicity?


The covid death breakdown is as follows: 61% White, 19% Hispanic, 15% Black, 4% Asian, 1% American Indian, and 0.2% Pacific Islander. If you compare these percentages with the general breakdown of the US as a country by race we find disparities in the Black and American Indian population. For reference: courtesy of the visual capitalist , we find that the US population is broken down as 60% white, 18.5% Hispanic, 12.2% black, 5.6% Asian, 0.7% American Indian, and 0.2% Pacific Islander. This is called the unweighted population percentage. If we weigh by the population demographics at the county level since the majority of cases occurred in cities where people are clustered and the demographics are different than if you take the whole country as a whole we find a slightly different story. We find disparities in white, black, and American Indian. Keep in mind that some deaths may be under-reported or misclassified.

From the CDC website the weighted population breakdown based on where COVID-19 infections were found shows the adjusted population breakdown to be 39.7% white, 34.4% Hispanic, 13.1% black, 10.1% Asian, 0.4% American Indian, and 0.2% Pacific Islander.

The CDC provides both estimates and helps explain the calculations so one can use whichever estimate supports their research. They do propose that if Covid-19 infections were assumed to be evenly widespread throughout the country then we wouldn't need to have a weighted population calculated.

Again, take any statistical inference with a critical perspective.

Lastly, what about deaths by age group?


Another concerning question is why so many elderly live in densely populated cities? They are most at risk and would benefit from living in smaller towns. Maybe it has do with social services and healthcare availability. That is another problem we will need to solve with an aging population. For now, clustering the elderly with the young during a pandemic is not the best option.

For those under 45 years old there is some comforting data point. There is a 97.5% chance a Covid-19 infection won't cause death for those under 45 years old. That is an interesting statistic for future pandemic economic policies and has implications for vaccine distribution plans. Note however, if you are under 45 years old and don't have insurance--consider being the first in line to get the vaccine. Seriously, just being treated for the flu let alone Covid-19 is too costly in this healthcare system compared to the cost of a vaccine.


Source of summary data: CDC