Have the protests started the second wave of Covid-19?
empirical evidence from the George Floyd rallies shows no increased transmission rates yet but next couple of weeks will be crucial as to how this unfolds.
Daily new positives for entire USA. Green dot marks George Floyd’s death (since 1 April)
Perhaps the most common restriction placed on societies throughout the world when the new coronavirus struck outside of China in March 2020 was the introduction of social distancing in public. People were expected to stand at least 1.5 m away from each other under the belief that few droplets from the mouth and nose would reach a distance beyond this. This policy was generally introduced for all public situations, whether outdoors or inside, despite a complete lack of evidence concerning any differences between the two in terms of infectiousness of the disease.
At the time it was the safe option as we knew little about it. As a consequence all gatherings of more than a few people were banned completely. All team sports were stopped throughout the world, religious gatherings halted, live music venues shut down, nightclubs closed. Even weddings and funerals were restricted in numbers. These regulations have caused monumental upheaval on billions of people’s lives, along with catastrophic economic impact on virtually all countries. Whilst I originally generally agreed with these constraints I now question if we are too slow in lifting some of them. In particular I think it is obvious now that the mechanics of indoor transmission are completely different to what happens outdoors. The virus does not appear to readily cause infection outdoors from admittedly limited data so far.
In hindsight this was suggested in March when Japanese researchers determined that secondary transmission of COVID-19 was 19 times more likely to occur inside than outside. Chinese researchers subsequently published that the vast majority of clusters in Wuhan in February occurred indoors or in transport, with only one from 318 outbreaks occurring in an outdoor environment. These studies occurred in winter and the disparity is likely to be even greater when it is summer or sunny and UV light from the sun has greater opportunity to deactivate the virus.
At one daily recent press briefing from the White House coronavirus task force, the acting undersecretary for science and technology at the Homeland Security Department, William N. Bryan, showed results of some studies by the agency at the U.S. Army’s high-level biosecurity laboratory. Here is a reproduction of his slide
A slide from Mr. Bryan’s presentation showing results from a recent lab report into half life of the coronavirus under different conditions.
This suggests that outside the virus is deactivated very rapidly both in aerosol form and on surfaces, particularly in summer. Similar numbers were published in the Journal of Infectious Diseases recently, where they suggest sunlight might be as effective as disinfectant in decontaminating surfaces.
Much of this is in pre-print form (not peer reviewed) and hence subject to the conditions of the experiments and may not be generalized so is open to criticism. However, the same could be said for near universal recommendations that social-distancing be maintained everywhere when there is not a shred of evidence that this substantially reduces outdoor risk for this particular virus, in particular when compared with universal mask-wearing, which is finally gaining traction in western societies.
This brings me to the main point of this article. Whilst the circumstances around the George Floyd protests are tragic, hopefully some good will come from them. Of course the best outcome would be a reduction (or elimination) of inherent racism in our societies but there may be other benefits too. Throughout the world people have been flaunting outdoor social distancing measures as they protest and rally for this cause. Health experts everywhere are up in arms about this suggesting it will lead to a huge increase in virus transmission. I am not so sure, based upon what I have read and summarized above about the low risk of outdoor transmission. Perhaps this is more in hope than reality but we will see.
It has been about 2 weeks now since the first rallies so I would expect to be seeing some increase in daily transmission rates by now in US counties which first had large demonstrations within a few days of his death (Minnesota, Washington DC, New York, Los Angeles and Atlanta). Whilst some demonstrators wore masks, it was nowhere near universal. In addition, use of tear gas, pepper sprays etc lead to increased coughing, sneezing etc which would also theoretically lead to greater transmission rates. Cheering, shouting and singing also increases risk. As some demonstrations also ended up in violence, many were incarcerated for a period of time. So if daily positive rates go up, how can we differentiate between transmission by being locked up in jail compared to being involved in a peaceful protest walk? Hopefully by contact tracing but this may take some time to sort out.
So far I am not seeing an increase in transmission rates in cities with early protests. Here are some graphs of daily transmission rates, which have also been smoothed to make trends easier to see. George Floyd’s death is marked by a green circle.
Minneapolis had protests the day after George Floyd died
Hennepin County was just over the peak of daily transmissions when George was murdered. In the 13 days since rates have continued to decline.
Protests did not start in Washington DC until 4 days after George’s death (May 29)
It is perhaps too early to see any changes in transmission rates due to protests in the District of Columbia. They are currently trending down. Since protests began this county has been the focal point of more recent rallies.
Protests erupted in Los Angeles May 27
In Los Angeles, daily cases were increasing until the time of his death, after which they have plateaued. Time will tell whether rates decrease or increase from this point.
Protests in New York City started May 29
New York was well over its peak by the time its rallies started. There is no indication of a new surge.
Atlanta had violent protests beginning on May 29
Atlanta has been more or less in a holding pattern over the last two months showing a slowly decreasing rate of daily transmissions. There is currently no sign of an upward trend.
I did make an attempt to compare US counties with and without rallies to see if there was any significant statistical difference between the two groups (paired T test if anyone is interested). Although I found no significant differences, I think such a test is fraught with difficulty as it is next to impossible to differentiate counties with/without rallies as they were so widespread and may have not have been reported accurately. Instead I decided just to look at the USA as a whole.
Daily new positives for entire USA. Green dot marks George Floyd’s death
The cyclic nature of this is due to a weekly, seven day cycle, with fewer reported cases on weekends. It is a sampling error and not a realistic trend.
Despite none of the largest centers of protest having a positive upward sloping trend , when we look at the USA as a whole there is the slightest positive slope in daily trends in the last two weeks since George’s death. Could this be caused by the demonstrations or is it just minor blip with the downward trend set to resume in a few days?
Originally when I first began to write this piece I was confident from prior research that the demonstrations would not affect the overall trends. I was hoping this could lead to a relaxation of social distancing outdoors. Now I am not so sure. Even if the overall trends do start to go up again in a so-called second wave, the main issue is that not all the protesters were outside all of the time. Many were incarcerated for a day or more. Jails are a hotbed of super spreading. Some people practiced social distancing whilst many others did not. Some wore masks, others did not. It would be almost impossible to make any conclusions about outdoor transmissions if an upward trend gains momentum. It may be necessary to look at cities outside of the USA with less violent demonstrations before any conclusions can be drawn.
The only thing that will show this is time. Perhaps this article is premature and a re-think is necessary in a couple of weeks when trends will be more obvious. For this reason I will update these graphs daily on this web blog but I hope this is food for thought for now.
Andrew Walsh PhD MD.
PS. All data is taken from here