This blog is a (much!) less-than-formal outlining of recent travels, events, happenings, thoughts and comments which tend to have some occupational relevance, but are on occasion nothing more than a means of passing the time while waiting for trains, planes & automobiles...

Thursday, 13 August 2020

Cherry Picked Information

Only 56 Million Cases to go until we reach H1N1 (Swine Flu) Cases from 2019!
Remind me again how long were schools shut down for?
What type of masks were popular back then?

Not having anything like as much data analysis to do as would be the case were we not in a pandemic, I've investigated the above meme which has appeared in my Facebook scroller due to posting by a Friend and getting liked/commented by enough other people to make it the first thing I saw.

The first thing that occurred to me was "is this data accurate"? I am not sure which data sources, location, or how recent it is, but as the poster has an intertest in US politics, it's probably about the US even though there are no citations to back up the statement.

However, it is consistent with the US CDC’s data. Over a period of one year (from April 12, 2009 to April 10, 2010), the CDC estimated there were 60.8 million cases of H1N1 in the United States [1]. Using the same source, the current cases of Covid-19 in the US are 4,920,369 (about 4.9 million) [2]. So as of today, we need roughly 55.9 million more Covid-19 cases (almost 56 million) to catch up on the 2009 H1N1 cases. So, yes, the data in the meme seems accurate.

Now, what about the argument? Although the conclusion is implied, the argument appears to be:
Premise 1: There were more 2009 H1N1 cases than the current pandemic
Premise 2: But we wear masks and close schools for this pandemic, but not for the 2009 pandemic.

Conclusion: we (or the US if my initial assumption is correct) are overreacting and don’t need to wear masks and close schools for this pandemic.

So on first reading, and without any sort or critical thought, it seems to make a sensible point, but a closer look shows it to be spurious. The argument rests on the hidden assumption that the case numbers are the primary determining factor for deciding if masks should be worn and schools should be closed.

But that doesn't make much sense as the average adult probably gets the common cold a few times a year, and given the US population, the cases of such a mild illness would be several hundred million a year. Clearly, other factors must be considered when deciding how to mitigate disease, whether that mitigation is with masks, closing schools, social distancing, or whatever else.

There are a number of important metrics in epidemiology such as incident rates, morbidity rates, mortality rates, hospitalisation rates, and many others, so only focusing on the cases is ridiculous, especially when these two diseases, H1N1 and Covid-19 are very different.

Going over each metric would be take way too long when there's cricket on the television, but we can clearly see that these two diseases are different by looking at the deaths in conjunction with the cases. The US coronavirus deaths to date attributed to COVID-19 is about 160,220 since January 2020, with an estimated 4.9 million cases.

However, H1N1 had substantially fewer (12,469 deaths from April 2009 to April 2010), from a whopping 60.8 million cases. But there is something else to think about. The death data for H1N1 is for one year, whereas the COVID-19 numbers are for only 8 months—and it may not slow down for several more months. Moreover, the H1N1 pandemic had relatively few precautions to attenuate its spread. Few people wore masks, there was little social distancing, etc.

So what does this mean? This means that over a period of 8 months, with social distancing, Covid-19 has more than 12 times the deaths (with social distancing and masks) from less than 1/12 the number of cases H1N1 had over one year. Though the 2009 pandemic was concerning, it was clearly nowhere near as deadly as this current pandemic. Therefore, at least concerning how society should react to the two diseases, it doesn’t make sense to compare these two diseases as equal.

So therefore this meme, which probably took seconds to compose, and maybe a couple of hours to pull apart with cited sources, makes a fallacious argument by using cherry-picked data from two very different diseases, and hyper focuses on case numbers as the sole metric for these decisions.

Who would have that sort thing would be all over T'Internet eh? :)


1. CDC website. 2009 H1N1 Pandemic (H1N1pdm09 virus). Last reviewed: June 11, 2019. Accessed Aug 12, 2020.

2. CDC website. Coronavirus Disease 2019 (COVID19). Cases, Data, and Surveillance. Last updated. Aug. 8, 2020, 07:00 PM. Accessed Aug 12, 2020.