During a recent call a normally sensible former professor, whom I respect deeply, informed me that Covid-19 was “yesterday’s story” in India. People were going about their business without worrying about it in the least. The government was keeping the Covid scare alive just to keep protesters off the streets. The disease was on its way out. The massive national vaccination drive was totally unnecessary and only being carried out to protect the investments of pharmaceutical companies.
As Covid conspiracy theories go this one wasn’t particularly wacky. Many in the United States had claimed that Covid-19 was a hoax. Kanye West, a pop singer who contested the US Presidential election in 2020, believes that vaccinations are “the mark of the beast”, intended to “put a chip inside us” and would ensure that many “can’t cross the gates of heaven”. West may seem delusional but still received over 60,000 votes.
Still this was rather unexpected, and somewhat disturbing, coming from a country which had suffered more than 150,000 confirmed — and probably many more unconfirmed- Covid-19 deaths.
The widespread confidence in India that the pandemic is over is reflected in articles like this. Anecdotal evidence suggests that behavioral fatigue has set in; fewer people are wearing masks, social distancing is not common and large public gatherings including public protests and big, fat Indian weddings are back. In a survey conducted in India in December, 69% of the respondents said that they were are in no hurry to be vaccinated against Covid-19. Rajiv Bajaj, an eminent industrialist, has declared on TV that he will not take a Covid-19 vaccine.
What is behind the mood in India? Is the pandemic really over for the country?
The Pandemic Game
At the simplest level the pandemic can be thought of as a game played between the SARS-CoV-2 virus (which causes Covid-19) and humans. The virus has a single objective; replication. It infects humans, creates millions of copies of itself within each infected human and these copies in turn infect other humans.
A relatively small percentage of Covid-19 infections lead to death or severe illness. The Case Fatality Rate (CFR) is the percentage of deaths among confirmed Covid-19 cases while the Infection Fatality Rate (IFR) is the percentage of deaths among all Covid-19 cases, confirmed or not.
Of course the IFR needs to be inferred from available data using statistical techniques since, we cannot know the number of unconfirmed cases. These numbers are usually estimated from serosurveys — sample surveys to determine what fraction of the population have antibodies triggered by a Covid-19 infection.
The IFR is not a constant of nature; unlike the charge of an electron or speed of light it does not remain invariant over time and space. It depends sharply on demography — older populations e.g. have higher IFR - and circumstances — the IFR is high when the heath care system is not good or gets overwhelmed. IFR varies considerably both between and within countries. A recent study of Indian data, estimates the Covid-19 IFR to be 0.12% for the state of Karnataka, 0.53% for the crowded megalopolis Mumbai and 5.64% among migrants in Bihar.
Covid-19 is a highly contagious disease. Therefore, even though Covid-19 leads to sickness and death of only a small percentage of the infected, the absolute number of hospitalizations and deaths is extremely high if the disease spreads freely. It follows that humanity’s objective in the pandemic game is to limit infection. From this perspective the pandemic is a zero sum game between humanity and the virus. The virus scores a point when it infects an individual; humanity scores when it prevents an infection.
There is a more subtle reason why we should try to limit transmission. Mutation is a well established virus strategy; it happens whenever an error occurs during replication. The more the number of infections, the greater the chance of virus mutation. Mutation gives rise to new variants. Some of these variants turn out to be more transmissible, i.e. better at infecting humans than the original variety. These infect more people and therefore cause more sickness and death, even if their IFR is no higher. Such more transmissible variants of SARS-CoV-2 seem to have arisen in the UK, South Africa and Brazil and are spreading rapidly around the world. Limiting transmission, apart from reducing illness and death, also limits opportunities for the virus to mutate.
The Return of the Zombie Idea
Yet not everyone agrees with this idea of the pandemic as a zero sum game. Some argue that humanity should take no special measure to control virus transmission, since such efforts are futile in the long run. While the virus spreads the young should be free to go about their business as usual since their IFR is low. Some care may be taken to protect the elderly and other more vulnerable segments. Once the proportion of infected in the population reaches the herd immunity threshold (HIT) the virus finds it increasingly hard to find fresh people to infect and the epidemic begins to die down. This approach is known as the herd immunity or focused protection strategy.
The allure of herd immunity is obvious. It seems to offer a “compassionate” path out of the pandemic without the economic and psychological pain that lockdowns inflict.
Unfortunately there is a lot of evidence that herd immunity does not work as an anti-Covid strategy.
It is virtually impossible to protect the elderly and vulnerable while letting the virus spread freely through the community. The following chart from an article by coronavirus celebrity Tomas Puyeo explains this vividly, better than words could.
The herd immunity strategy leads to more cases and more cases in turn lead to more deaths. UK, Sweden and the Netherlands which have all flirted with herd immunity strategy have paid for that with appallingly high Covid-19 mortality rates.
It is often claimed that it is unfair to compare India with with such countries.
It seems plausible that herd immunity might work better for India given its much younger population.
Unfortunately there is the disturbing counter-example of Manaus, a city in Brazil which serves as the gateway to the Amazon forest.
Manaus suffered a terrible Covid-19 outbreak in April 2020. The healthcare system was overwhelmed and mass graves had to be dug.
Antibody testing among samples of blood donors suggested that 76% of the population of Manaus had been infected by SARS-Cov-2 by October. That was above almost all estimates of the HIT for Covid-19. It was expected therefore that herd immunity would protect Manaus from a second Covid-19 wave.
It didn’t. The case and death curve did flatten in Manaus after the April peak and stayed flat till November. However in January the second wave struck Manaus with terrifying force. The cemetery had to be extended again and hospitals ran out of oxygen.
What accounts for Manaus’ second wave? The jury is still out. It is possible that the antibody survey of blood donors overestimated the percentage infected but this is unlikely; mandatory exclusion of donors with Covid-19 symptoms would more probably have led to an underestimate. The immunity conferred by antibodies may have waned by December. Most plausibly, it could be due to a newer variant which is more transmissible and against which the antibodies generated by the previous infection don’t work well. Regardless of the reason, Manaus illustrates how dangerous relying on herd immunity could be.
From an Indian perspective, the most important fact about Manaus is that it has a fairly young population quite similar to India’s. In spite of its young demographic, Manaus’ Covid-19 mortality rate is over 1,200 per million; extrapolated to India this would mean over 1.5 million deaths.
Yet, in spite of all its proven dangers, the herd immunity strategy keeps resurfacing to such an extent that it has been dubbed the “zombie idea of the pandemic”. And now herd immunity is back in India with a vengeance.
Why is the Indian zombie on the prowl again?
Taming the Supermodel
The herd immunity strategy has had many distinguished proponents among Indians and persons of Indian origin abroad. These include the industrialist Rajiv Bajaj, the economist Debraj Ray, the statistician Jay Bhattacharya, the epidemiologist Sunetra Gupta and the scientist Partha Pratim Majumdar. Indians heard a lot about herd immunity during the early days of the pandemic but much less during the terrible summer of 2020 when Covid-19 patients in Mumbai, Delhi and Ahmedabad struggled desperately to find hospital beds.
However, both Covid-19 cases and deaths peaked in September and have been going down ever since. Contrary to expectations, election rallies in Bihar and farmer protests near Delhi, involving thousands of people packed together and not wearing masks, apparently did not trigger a second wave. Indians have by now had time to forget what happened last year and convince themselves that the worst is behind them.
The reason for decrease in India’s caseload is not fully understood and is likely due to a complex combination of factors. However, herd immunity seems to offer an easy explanation. Since instant explanations are always popular it is no surprise that many are claiming that cases are decreasing in India because the country has crossed or is fast approaching the HIT.
The Covid-19 Indian national supermodel, authored by 3 extremely distinguished Indian scientists, Manindra Agarwal, Mathukumalli Vidyasagar and Madhuri Kanetkar, and reported widely in the Indian press, offers support for this view. According to the first author, the “supermodel” estimates that about 30% of the total population were infected by Covid-19 at the peak of the outbreak in September and predicted the number to go up to 50% by February. Others think that the figure may be even higher.
Mathematical models are indispensable and impact almost every aspect of our lives. Trying to navigate an epidemic without models is like trying to steer a ship at sea without charts or navigational instruments. Yet these powerful beasts must be handled with care; otherwise they can cause great damage.
During the pandemic we have seen the tragic consequence of ignoring scientists in countries such as Brazil and the United States. However, the United Kingdom illustrates the opposite error of policy makers blindly relying on mathematical models which they don’t understand. The UK kept its borders open and imposed no testing and quarantine restrictions on visitors for several months at the start of the pandemic, because it trusted “scientific advice” that measures to screen arrivals would have “negligible impact” on virus spread.
The Indian supermodel, like all models, relies on data and assumptions to arrive at conclusions. Like most models it too has limitations. These have been discussed lucidly by the noted epidemiologist Gautam Menon.
The third Indian national serosurvey, conducted in January 2021, estimates that 21.5% of the Indian population have been infected by Covid-19. This is far larger than the number of confirmed cases (which is less than 1% of the total population) but well below the supermodel’s prediction. Of course serosurveys have their limitations too; the results may be inaccurate either because of sampling errors or the antibody test used. Moreover, the country-wide infection rate hides large variations. Densely populated slums in large cities such as Mumbai and Delhi may very well have achieved herd immunity. Nevertheless, it is important for citizens and the government to continue to be cautious and act on the assumption that the country, as a whole, is far from herd immunity and the majority remain susceptible.
The Hygiene Hypothesis
Many Indians —both lay people and medical doctors- believe that prior exposure to germs confers on Indians a degree of immunity to Covid-19. This belief has a formal name in medical research; it is called the Hygiene Hypothesis.
This intriguing and plausible idea is explored in a recent paper by Bithika Chatterjee, Rajeeva Karandikar and Shekhar Mande. This deserves attention because Mande and Karandikar both have very distinguished track records; both have won the Bhatnagar Award -India’s highest scientific honor — and Karandikar is undoubtedly a mathematician of the highest order.
The starting point of the paper is the undoubtedly correct observation that richer countries have tended to have much higher Covid-19 mortality rates than poorer one.
India’s Covid-19 mortality rates for example are much lower than those of most European and American countries, even if we allow for considerable under-reporting of deaths. Quite naturally, the Indian government has been quick to cite these figures as evidence of its success in managing the disease.
India’s performance is less impressive, though, if we look east.
What accounts for the difference in Covid-19 mortality rates between countries?
Chatterjee et al. employ the well-known technique of linear regression to answer this question. They fit regression lines that plot the logarithms of the mortality rates (in deaths per million) against several combinations of possible explanatory variables.
Their analysis shows that Covid-19 mortality is best explained by 3 groups of variables; demographics (e.g. proportion of population above 65); sanitation (e.g. availability of safe drinking water); and prevalence of autoimmune diseases like Type 1 Diabetes Mellitus.
Covid-19 Mortality correlates positively with older and more urban populations. It correlates negatively with improved sanitation; of course poor sanitation is also a likely a cause of germ exposure. Finally Covid-19 mortality correlates positively with prevalence of autoimmune disorders; it is believed that autoimmune diseases may be caused by insufficient prior exposure to germs.
How compelling is the evidence?
The explanatory variables considered by the author are highly correlated. After all richer countries tend to have older and more urban populations, better sanitation and higher levels of autoimmune diseases. This makes it hard to assess the impact of each individual variable on mortality. Technically this issue is known as multicollinearity; it is not clear how the authors address it.
The paper presents statistical correlations which are far from universal. Several rich countries of the Asia-Pacific region (Japan, South Korea, Singapore, Taiwan, Australia, New Zealand) probably enjoy better sanitation standards than poorer countries such as India and yet have lower Covid-19 mortality rates. It is not clear how this can be squared with the Hygiene Hypothesis.
Finally, the Hygiene Hypothesis works at the level of individuals. To test it we need to compare the incidence of Covid-19 among individuals with different degrees of prior germ exposure. Arguing that country-level data establishes the Hygiene Hypothesis is surely a leap of faith.
Some claim that the low mortality rate in Dharavai, Mumbai and Asia’s largest slum, supports the hygiene hypothesis. This is wrong. Dharavi’s mortality rate would undoubtedly have been much higher in the absence or aggressive public health measures. In fact one major limitation of the paper by Chatterjee et al. is that it does not factor in the effectiveness of the response by each country to the pandemic, even though that is probably the most important determinant of the mortality rate.
At present the Hygiene Hypothesis remains an intriguing but unproven conjecture. It merits further investigation but neither Indian citizens nor governments (central and state) should rely on innate immunity to protect them from Covid-19.
The Vaccination Endgame
The hygiene hypothesis and belief in herd immunity are complementary. They suggest that a large part of the Indian population is protected against Covid-19 because of immunity - either innate or acquired through prior infection. Both have gained considerable traction in India and could lull Indians into a sense of complacency. Unfortunately that could have fatal consequences.
Proponents of herd immunity strategy often present it as the only alternative to lockdowns. Nothing could be further from truth. Many countries such as Australia, Singapore, Taiwan, South Korea and Vietnam have proved that it is possible for societies to keep Covid cases down and the economy functioning by adhering to well established public health measures. Some like contact tracing, worked well for the state of Kerala at the beginning of the pandemic and should work for India now that the caseload is low. Lockdowns should only be used as a last resort and then preferably be short, sharp and localized.
And above all India needs to take advantage of the lull to vaccinate as many people as it can.
With Covid-19 vaccinations the pandemic endgame has begun. Humanity will win.
India, dubbed the “pharmacy of the world”, is very well positioned for the endgame. It manufactures about 70% of world’s vaccines. Serum Institute of India is the world’s largest vaccine manufacturer.
Two Covid-19 vaccines are currently manufactured in India. The first — branded as Covishield in India — was developed by the University of Oxford and Astrazenica and is manufactured by the Serum Institute. The other -Covaxin — is manufactured by Bharat Biotech, an Indian company which developed it in collaboration with the Indian Council of Medical Research (ICMR). Several more vaccines are planned.
India’s has begun vaccinating its healthcare workers. The effort is massive and well planned. However, it did make a completely avoidable slip which could derail its vaccination effort.
The final phases of vaccine development involve human trials. Phase 1 and Phase 2 clinical trials of a vaccine are conducted on relatively small samples (typically a few hundred) to test whether the vaccine is safe and immunogenic (elicits an immune response). During the Phase 3 trials the vaccine and a placebo are administered to a very large number of people, with neither the participants not the administrators knowing who has been administered which. This is very important to determine the efficacy of the vaccine.
For example, 43,458 volunteers were enrolled in the Pfizer Biontech Covid-19 vaccine (BNT162b2) Phase 3 trial with 21,720 in the vaccine and 21,728 in the placebo groups. The trial was stopped when 170 of the participants contracted the disease. 162 of them were in the placebo and 8 in the vaccine group. The infection rate was therefore about 0.74% in the placebo and 0.036% in the vaccine group. Reduction in risk of disease or vaccine efficacy is (0.74–.036)/0.74 or ~95%.
India authorized the public use of both Covishield and Covaxin in its vaccination program. The results of the Oxford-Astrazenica vaccine Phase 3 clinical trials in Brazil, South Africa and the UK, have been published. Phase 1 and Phase 2 trials of Covaxin have reported encouraging results about its safety and immunogenicity. However, the Covaxin Phase 3 clinical trials are ongoing and so its efficacy is currently unknown.
Vaccine hesitancy increased with the use of a vaccine of unproven efficacy, especially since recipients are not allowed to choose the vaccine that they would receive. Doctors from a number of institutions have requested to be vaccinated with Covishield. At least one Indian state has informed the central government that it will not use Covaxin till Phase 3 trial data is available.
The government reacted angrily. In a letter sent to Chief Secretaries of all Indian states, Ajay Bhalla, Home Secretary of the Indian Central Government, darkly hinted that vaccine hesitancy was due to “ unfounded scare mongering” by “vested interests” and asked them to take penal action against everyone spreading “such ill-informed rumours”. Unfortunately trust in vaccines cannot be established by coercion. Penal measures would probably be counter productive and just increase the distrust. Transparency, education and persuasion are the only tools for overcoming vaccine hesitancy.
India was of course not the first country to authorize the public use of a vaccine which had not completed its Phase 3 trial. Russia and China had both done so. However, in India’s case the reason for pushing Covaxin at this stage is not clear. The Serum Institute currently manufactures about 2 million doses of Covishield each day and is rapidly ramping up capacity. India is currently vaccinating around 350,000 people each day. Obviously capacity will need to be ramped up considerably but Covaxin Phase 3 trial results are expected well before daily requirements of vaccine exceed Covishield manufacturing capacity.
The world should be grateful for Serum Institute’s surplus production capacity. Covishield is being gifted or sold to many nations including Afghanistan, Bangladesh, Brazil, Myanmar, Morocco, Nepal, Seychelles and South Africa.
Bharat Biotech has a strong track record. Covaxin is developed using an inactivated virus — a tried and tested technique. So there is good reason to hope that the clinical trial will establish satisfactory Covaxin efficacy and vaccine hesitancy will decline. However, this is a risky gambit. If Covaxin efficacy turn out to be low that will deal a great blow to the credibility of India’s vaccination program.
Unfortunately the pandemic endgame will be long. The virus has upped its game with new variants. Bad moves will delay victory and result in more deaths & hospitalization on the way.
Disclosure: I am the Director of Smart Consulting Solutions Pte Ltd, incorporated in Singapore and its subsidiary Radix Analytics Pvt Ltd, incorporated in India. I am also a Visiting Faculty Member at the Indian Institute of Management, Udaipur. The opinions expressed in this article are solely mine and not necessarily shared by any company or institution with which I am affiliated.
Prof. Gautam Menon’s Facebook posts are a treasure trove of information regarding Covid-19. One of my arguments against the Hygiene Hypothesis is a paraphrase of one of his remarks in a webinar, though he is of course not responsible for my views.
I am, as always, grateful to my friend Dr. Ashish Kumar Dawn for his insightful comments and especially for introducing me to the Hygiene Hypothesis though he doesn’t fully share my views about this.