Analysing data from Ahmedabad, Dr Dileep Mavalankar, Director, Indian Institute of Public Health, Gandhinagar says it supports a theory by Prof. Karl Friston of University College London, and Prof. Britton of Sweden and others, that in the immunity of infectious diseases like COVID-19 there is some “Dark Matter” – which is not easily measurable but can be modeled based on the data to show that some countries will see herd immunity-like effects at much lower levels of infection in the population. The effects of herd immunity may be seen even at 20 to 40 per cent infection rates.
We are told that herd immunity or vaccine are the two ways the COVID-19 epidemic will go down in the world. Herd immunity needs 70 per cent plus people to be infected and immune, which is very high order and may lead to millions of deaths at the current mortality rates.
Vaccine is months away if one of the 180 groups succeed to produce an effective one. India has not been able to vaccinate all children with basic immunisations which are available for last 20 years. So expecting that a new vaccine will be given to 70 per cent of Indians to build herd immunity is nothing short of miracle.
So we have only one option of herd immunity via natural infection. It is also not easy to imagine that 70 per cent of Indians will get infected soon. The currently reported cases – about 1,000,000 is 0.077 per cent of the Indian population. If we assume that actual infection rate is 100 more than reported numbers, then also it comes to 7.7 per cent of India’s population to be infected – far cry from 70 per cent needed for herd immunity.
The theory of “Dark Matter” in immunity modelling of infectious diseases
Fortunately, recently some brilliant scientists have developed a new idea using European data to show that in the immunity of infectious diseases like COVID-19 there is some “Dark Matter” – which is not easily measurable but can be modeled based on the data to show that some countries will see herd immunity like effects at much lower levels of infection in the population. The effects of herd immunity may be seen even at 20 to 40 per cent infection rates.
This has been argued by Prof. Karl Friston of University College London, and Prof. Britton of Sweden and others. They argue and we support them that in a theoretical population where all are equally susceptible one needs 70 per cent immunity to achieve herd immunity effect.
But in COVID-19 like real epidemics even with a new virus – not all in any community are susceptible to the disease. Some are highly susceptible and others have a graded level of resistance to the disease due to various reasons – cross immunity due to past infections by corona viruses, or other similar viruses, some other infections or immunisations etc.
Secondly, 70 per cent immunity is required when we assume that all people in the population are randomly interacting with each other – this only happens in a simple mathematical model and not in real society anywhere.
In real society the interaction is very limited due to geography, cast, class, gender, age etc. And hence the chance of transmission becomes much less than theoretical random interaction. This also lowers the per cent of immunity needed to achieve herd immunity in real world. Prof. Friston calls this interactive population as effective population – rest of the population is not interacting and sequestrated.
For example, my mother who is 90 years and bedridden interacts with only three people and unless one of these three are infected and can pass her the infection she is practically not exposed, sequestrated or not in the play to get infection. As per Friston, 50 per cent population is such sequestrated or not exposed population.
Then of the rest, another 50 per cent (or 25 per cent of the total) are not susceptible due to past infection due to similar other virus or other immunogenic factors, or they may have natural resistance to COVID-19. It is not necessary that because COVID-19 is a new virus all must be susceptible to it.
In 2006, India and the world was ravaged by a re-emerging mosquito borne virus called chikungunya, a virus that returned in new-and-improved form after 40 years and hence one can assume that no one would have immunity to that virus. But in that epidemic as well, only 14 million cases were reported in India, out of 1 billion plus people. Ahmedabad had a very bad epidemic with 4000 additional deaths in three months of the epidemic, but still only a small fraction of the population was infected. The same thing happened in the recent H1N1 epidemic – very few were infected even though the virus had mutated.
Our belief that many people are resistant to this virus in many countries of the world is further supported by low levels of secondary attack rate in close family contacts. Before a case of COVID-19 is tested and diagnosed, generally he or she is symptomatic with cough, cold and fever for few days – during that time the case has ample opportunity to pass on the infection to the close family members as there is no social distancing or wearing mask at home.
But the studies of SARS-CoV-2 virus in a family set up generally show that only 10-20 per cent of close contacts of a proven COVID-19 case develops the infection – that means that 80-90 per cent of very close family contacts of the case of COVID-19 do not develop the disease – it shows that they are resistant to the infection in spite of heavy exposure for several days from the case, before isolation of the case.
The other explanation for this non-spreading of the disease in homes could be that most of the clinical cases are not infectious due to low viral load or other inhibitory factors. Further research is needed on this. The secondary attack rate among spouses is also only 45 per cent and not 100 per cent even after sleeping in the same bed with an infected person!! So it seems clear that substantial proportion of the population in a real country or city are not exposed or not susceptible to this new disease.
After removing not exposed and not susceptible population about 25 per cent of population remains that is susceptible to COVID-19 and can get infected and transmit the disease. So if we take 70 per cent of 25 per cent it comes to 17 per cent. So in any area if about 17 per cent get infected and become immune we should see effect of herd immunity.
Empirical proof of the theory?
This is all theory and modeling. Is there empirical proof that this is happening in India??
Ahmedabad with a population of 6 million has reported about 22,000 cases so far (till July 15) and the epidemic is declining. Looking at the epidemic curve, we see that there may be a total of say 30,000 cases over next month or so. This means that the epidemic has ended/will end in one month with about 0.5 per cent of the population reporting infection on RT PCR test. Why would an epidemic end in Ahmedabad with only half per cent of population showing infection, especially after lockdown is lifted and people have started to mingle with each other??
The only explanation is that there is some form of herd immunity, plus non-susceptible population plus not exposed population which blocks the further spread of the infection in the city in spite of opening of lock down.
The other fact to be kept in mind in COVID-19 is that many studies in the world have shown that there are about 20-50 times more infection in the community than the diagnosed cases. So if we multiple 0.5 per cent reported cases by 20-50 we get population per cent infected as 10-25 per cent. And this is the range that Friston et al have indicated for herd immunity to show up in a real community.
So what we are seeing (see graph: Daily new COVID-19 cases in city of Ahmedabad, March to July 2020) in Ahmedabad city, with a population of 6 million, with about 30,000 cases is reaching of herd immunity taking into effect the immunological “Dark Matter” which Friston has described.
Daily new COVID-19 cases in city of Ahmedabad, March to July 2020
(These are the personal views of the author)