The TLDR: A serological survey found that 51% of residents in parts of Pune have been infected with the virus. But is this good or bad news? New research suggests that herd immunity can kick in if 50% (or even less) of a given population is infected—which in turn can stop the spread of the disease. Translation: we may not have to wait for the vaccine. But there’s one big problem: the virus does not spread evenly within nations—or even cities. We took a closer look at herd immunity in the light of the Pune survey.

Tell me about the survey, first

The serological survey tests for presence of antibodies—which indicate that the person was infected sometime in the past two weeks (we explained how the survey works here). There have been similar surveys in Delhi, Mumbai, Ahmedabad etc. 

• The Pune survey is the most recent, conducted between July 20 and August 5. 
• 1,664 persons from five city sub-wards with a high number of cases were tested—and 51% were found to be infected.
• None of them had reported symptoms or ever been tested for Covid-19—which means that they had been infected and recovered without their knowledge.
• The survey uncovered an income gap: 62% of residents in slum hutments were infected compared to 43% of those who live in bungalows.  
• Also: the rate was 60% among those living in the smallest homes (less than 150 sqft) and 34% among those in the largest (above 501 sqft).

What we learned: The disease has spread far more extensively than the official numbers indicate. Pune’s recorded number of cases is 132,000.

Compared to other cities: Pune’s numbers are on the high side. Delhi’s survey revealed 23% were infected. Mumbai recorded 57% in slums and 16% in buildings. The overall number in Ahmedabad: 47%. But since Pune is the most recent to be surveyed, other cities which were surveyed earlier may well have similar numbers now.

Ok, how is this good news?

Because it suggests that at least some parts of the country are close to achieving ‘herd immunity’. This happens when enough people in a given community/population are infected. They develop antibodies and become immune to the disease. This makes it harder for the virus to jump from one person to the next.

The R0 value: of a virus indicates how many people a virus can potentially infect. The higher the number (pronounced R-naught), the more likely it is to spread. But when more people become immune, that number goes down. So let’s say the virus’ R0 factor is around 2.5 (though estimates vary widely). Now if that person encounters ten uninfected people, then they will spread the disease to two others. But if say four of those ten are immune, then it will only spread to one person. 

Point to note: This is also why vaccines work. They artificially create conditions of herd immunity—as illustrated by the visualisation below.

Also: This doesn’t mean that there won’t be new infections, but the spread will rapidly recede. Or as one scientist puts it: “It doesn’t mean you won’t be able to start a fire at all, but that outbreak is going to die”.

So 50% is enough?

At the outset of the pandemic, medical consensus indicated that at least 70% of the population has to be infected for herd immunity to kick in. But new research—based on statistical modelling, mind you—indicates the threshold may be lower. The estimates vary from 50% to as low as 10-20%. According to one of the mathematicians who puts the number at 43%:

“Assuming the virus ferrets out the most outgoing and most susceptible in the first wave, immunity following a wave of infection is distributed more efficiently than with a vaccination campaign that seeks to protect everyone.”

Quanta has a great visualisation of this:

Sounds kinda optimistic…

As we said, these are mathematical models—which don’t often work as well in practice. 

One big problem: is that the infection has not spread evenly within countries—or even cities. 

• As we noted, Mumbai is an excellent example: 57% in slums vs 16% in buildings. 
• In New York, there’s a similarly large gap: 68% at a clinic in Queens compared to just 13% at a Brooklyn facility. 
• Herd immunity favours the less affluent. Those crammed into smaller spaces are more likely to catch the disease, and hence develop herd immunity.
• So the very term becomes meaningless when there are such wide disparities within a small geographical area—and seamless interaction between its residents.

The other problem: We still don’t know how long immunity lasts within infected people—though the latest research shows that both antibodies and disease-fighting cells stay in the body for months after infection. Even so, doctors are cautious: “Protection against reinfection cannot be fully confirmed until there is proof that most people who encounter the virus a second time are actually able to keep it at bay.”

Last, but not least: As a leading virologist tells The Hindustan Times: “Why is it that if 51% of the population was infected… that we keep reporting more number of cases? Clearly, antibodies do not guarantee herd immunity.”  

Point to note: Pune now has the maximum number of cases in Maharashtra—and is #2 in the country. Also: 2,800 new cases were recorded on Sunday—compared to Mumbai whose daily count is between 1-2K. (Note: the Pune number has been corrected.)

So what good is any of this?

A higher percentage of infections can still be modestly good news—even if it doesn’t confer herd immunity. One, it can minimise the effect of a second wave in places that bore the first brunt:

“Definitely the disease would not spread as well if it gets back into New York… The same level of behavior change will have more effect on the disease now than it did four months ago.”

Two, it offers greater guidance on what areas and demographics are most vulnerable—and suggests who needs to be vaccinated first.

Reading list

The Telegraph has the most details on the Pune survey. New York Times digs into the latest research on herd immunity. Worth your time: Quanta’s deep dive into the “tricky math” of herd immunity for Covid-19. Boston.com has the latest research on immune responses to infection.