2020 and the rise of half-baked science

 2020 and the rise of half-baked science

 Poor science goes viral

Apparently, over 50% of the populations in some countries took to baking during the COVID-19 lockdowns in 2020. (Baking out of boredom: Home baking category report 2020 | Category Report | The Grocer) So I think it’s only appropriate that baking should figure in anything I write about the year. Also, “half-baked” seems like a nice metaphor for a lot of what was put forth in the name of science. And by science I mean medical science, specifically related to COVID-19, of which I can claim to know a bit about. Of course, good science has helped us get effective drugs and vaccines over an unimaginably short time-span. But on the other hand, thanks to our collective desperation in the face of the pandemic, there was an all-round lowering of the threshold for BS. And poor science went viral.

 

Flawed analyses and false promises

First, there were the predictions, most of which were churned out by people who had a rudimentary understanding of disease modelling and little or no understanding of the disease. (http://randomramblings2018.blogspot.com/2020/10/pandemic-predictions.html) Many of them projected an aura of smug certainty, but were sensationally off the mark. But who was keeping count? Even if called out, there were a hundred reasons that could be invoked to explain why the “model” hadn’t worked (lockdown was too strict/too lax, not enough testing/too much testing, or anything else you could come up with). If there was an award for the most dramatically certain (and of course completely wrong) prediction, it must go to this gem, COVID19PredictionPaper20200426.pdf (flasog.org). The author very helpfully provided exact dates for when the pandemic would end for every country in the world!

 

And then there were people who confidently asserted that air temperature had a “massive effect” on COVID-19 transmission and recommended that “all indoor air conditioning systems be operated at or above 25 degree Celsius”! (https://dx.doi.org/10.2139/ssrn.3575404) Others clamoured for “mass administration of vitamin D supplements” to the entire population as they believed that lower vitamin D levels in blood meant a higher risk of dying from COVID-19. Some of these conclusions were drawn from correlating country-level average vitamin D levels, and deaths due to COVID-19 at the country level. (https://doi.org/10.1007/s40520-020-01570-8) These studies ignored something called the ecological fallacy that comes into play when drawing conclusions about individuals from aggregate data. (Please see http://randomramblings2018.blogspot.com/2019/06/eggs-meat-and-happiness.html for another fallacious analysis). Those interested may also want to read about how the stork population in Brandenburg, Germany, correlated rather nicely with the number of babies born! (New evidence for the theory of the stork! https://doi.org/10.1111/j.1365-3016.2003.00534.x) Advocates of using Zinc for COVID-19 based their claims on even flimsier evidence: observations from in vitro studies in the laboratory. And I have previously lamented about inflated claims about a bunch of drugs for treating COVID-19 that have little or no supporting data. (http://randomramblings2018.blogspot.com/2020/06/home-remedies-hydroxychloroquine-and.html) The sad result of all the frantic publishing has been that for much of the duration of the pandemic, many hospitals and doctors around the world have been prescribing a cocktail of useless (and not entirely harmless) medicines, and some still continue to do so.

 

Unfiltered publishing and spotty research

To my mind, the most important reason for the spread of BS during the pandemic is the ease with which anyone can disseminate their work, with practically no quality checks. The ubiquitous pre-print servers offer a low barrier for publication without any peer-review, and social media offers unlimited publicity. Until recently, these were not popular with medical researchers. But the pandemic changed all that. Monthly uploads to medRxiv (medarchive) rose many fold, to several hundred thousand in 2020. But, a recent analysis showed that just about 15% of the uploaded COVID-19 related papers were subsequently published in peer-reviewed medical journals. (https://doi.org/10.1101/2020.09.04.20188771) This essentially means that the 85% of the content which didn’t make the cut, is still out there, miseducating doctors, and more dangerously, the lay public. It’s no wonder that we saw television anchors graduate to becoming self-proclaimed epidemiologists and health experts in 2020. It was just a matter of time that newsworthiness became the yardstick by which to assess the importance of scientific work. The consequences of this shift can range from being merely amusing to potentially dangerous. A recent analysis touting the “hygiene hypothesis” as the reason for fewer COVID-19 related deaths in countries in Asia and Africa, was lapped up by the media. (https://doi.org/10.1101/2020.07.31.20165696) The study claimed that poor hygiene (as measured by lack of access to clean water and sanitation) protected people in these countries from COVID-19 because they were used to getting infected throughout their lives. This study checks all the boxes for poor quality (invalid input data, inappropriate analysis and interpretation) and wouldn’t have cleared peer-review in a decent medical journal. But then, here it was, in the limelight, pushing a half-baked theory, misinforming the lay public, and discomfiting health professionals.

 

Now, don’t get me wrong. I am all for democratization (the intent of the creators of pre-print servers was to wrest control of publishing from the big companies and empower authors). But putting unvetted material up for public consumption is a sure recipe for muddying the waters. Hopefully, we will learn from the lessons of 2020; researchers will be more restrained and publishers will exercise some judgement. Or maybe it’s just wishful thinking on my part. Few can resist the lure of social media celebrity! Anyway, I say all this only in the context of medicine. Theoretical physicists can go publish their findings directly on twitter if they wish to!