COVID-19 Day 60: Herd Immunity Could Be Smaller Than We Thought


© Jens Lucking Photography

It’s been 60 days since the start of the National Emergency and even longer since my family has been in COVID home-isolation in San Francisco. I was planning on making this 60th-day anniversary post about a topic other than COVID-19, as I’ve grown a bit fatigued on the subject. But I learned about something new the other night that changed my mind.

A two scientific papers came out last week that independently share a similar implication: the threshold “herd immunity” is much lower than previously thought. It can’t be stressed enough that these papers are still under peer-review, so that their equations and concepts still need to be validated by other experts in statistics and epidemiology. But more importantly, drawing too far a conclusion could be politically explosive, especially in our current climate.

‘Herd immunity’ is a concept in vaccination and epidemiology where you only need a part of the ‘herd’ to be immune to make it improbable for a pathogen to continue to spread. This percentage is a ratio derived from the rate of reproductive spread of a virus (expressed as Rø or “R-naught”). Given how quickly COVID-19 has spread, Johns Hopkins epidemiologists have estimated that 70% of the US population would need to be immune from COVID-19 to reach herd immunity. As discussed in my previous article, my unwashed calculations estimated that would result in 240,000 deaths, or as many as 2.1M estimated by the Imperial College in early March.

‘Super-Spreaders’ was a term coined by scientists investigating the SARS and MERS outbreaks. These are events, gatherings, or individuals that have significantly higher than average ability to spread a disease, creating clusters of infections. For COVID-19 a number of super-spreader events have been documented including the Biogen medical conference in Boston, the CES electronics convention in Nevada, and cruise shipsThe 2015 MERS outbreak in South Korea can be traced to one super-spreader individual, a business traveler who spread the disease to numerous individuals, two of whom became Super-Spreaders themselves in their communities.  

Two papers, posted independently of each other, sought to refine the “classic” herd immunity ratio by incorporating super-spreader effects. Rø is traditionally estimated using an average member of a population. But in the real world, a disease spread can be highly influenced by super-spreaders. More importantly by eliminating super-spreaders from the equation, after they become immune, lowering the Rø value faster. 

The May 6 paper “The disease-induced herd immunity level for Covid-19 is substantially lower than the classical herd immunity levelestimates that depending on the variability of Rø, the percentage to reach herd immunity isn’t 70% but falls to 43%-60%.

The May 2 paper “Individual variation in susceptibility or exposure to SARS-CoV-2 lowers the herd immunity threshold” estimates that the threshold for herd immunity could be below 20%! 

A third unrelated paper from Google data scientists, coincidentally released the same week as those other two papers, offers supporting insight. It too incorporates the super-spreader effect for a different problem; how to efficiently employ antibody testing in the general population.

It bears repeating that these papers are essentially working-drafts. They haven’t been fully peer-reviewed by other experts in statistics and epidemiology. Science is the process of independently testing hypotheses to gain greater certainty. And these papers are statistical models not experimental findings. These new models which, if validated, are intended to be used for future clinical or epidemiological experiments. 

As we saw with other statical models, such as the Imperial College paper, the accuracy of results depends on the data that goes into it. And even if the data is accurate, the results produced are a range of possibilities that have to be interpreted. The Imperial College model offered a range of possible death tolls based on different strategies employed by the government, to slow the spread of COVID-19. But the advisors chose to stress the worst-case scenario numbers to influence the UK Prime Minister and the President of the US.

It’s easy to see how a theoretical model that implies, ‘herd immunity could require only 10% of the population’ could be easily rephrased by gullible or biased journalists into a headlines claiming certainty that, ‘we have reached herd immunity already.’ This would be ammunition for various political and ideological factions with their own agenda. We’ve already seen speculative “science” claims from far-left anti-vaxxers and far-right anti-government groups in reopen protests.

Furthermore, the Rø and the herd immunity number in the real world is governed by more variables than just super-spreaders. Population density, demographics, compliance with safety-protocols, public-transportation, and many more. We can see the results of these different variables in the vastly different infection and death numbers between each of the states. A nuanced approach is obviously required.

I debated sharing this info myself. But I offer them here as a means of vaccinating the reader to the politicization of science. To be further transparent, I believe the wholesale self-quarantine of the entire population due to COVID-19 will be found to be largely counter-productive. Because of that, the national economic shutdown did more harm than good, based on what we now know about COVID-19’s Rø and it’s chance of mortal outcomes. But my personal insights, as well as those described in these papers, were gleaned the hard way. As a nation, we’ve had over 80,000 deaths from COVID-19 and associated causes, despite our best efforts.

We have been asked to exercise a higher level of caution to stop the spread of this pandemic; wash our hands, wear face-masks, and keep a safe distance from one another. I ask you to exercise a similar level of caution in drawing and sharing your conclusions about this information.


Comments are closed.