Using epidemiological models, Dr Mukherjee's team came up with real-time situational analyses to help policymakers negotiate the pandemic better.
One of the earliest papers on the impact of social distancing in mitigating the chances of contracting Covid-19 came from a team of researchers led by Bhramar Mukherjee, professor and chair of biostatistics at the University of Michigan in the United States.
Using epidemiological models, Dr Mukherjee’s team came up with real-time situational analyses to help policymakers negotiate the pandemic better. In an interview with The Indian Express, Dr Mukherjee spoke of the inferences that could be drawn from Delhi’s serological survey, what could be behind the falling numbers in Delhi, and why women leaders seemed to be handling the crisis better.
Could you tell us how you arrived at the initial projections about the spread of the virus in India?
We started this work on March 16, the first day of the stay-at-home order in Michigan. We never imagined this work would have such significant impact. It started as an altruistic mission by a group of quarantined scientists, confined in each of their locations and worried about their loved ones in India. We wanted to repurpose our anxiety to something meaningful. Initially, I borrowed the modelling tool that my colleague Peter Song and his team developed for analysing the Wuhan data. Then, my team worked on multiple modifications to incorporate the context-specific nuance for India. One major innovation was to create an app that updates projections each day and produces a daily summary dashboard, relevant for policymakers.
New Delhi’s serological survey report shows that 23.5 per cent people have already been exposed to the virus. Given the comparably low fatality rate, is that a hopeful sign?
Let me try to itemise the good and bad through a forensic examination of the data:
If the Delhi serosurvey is reliable we have had 4.7-5 million past infections already in Delhi, of which we have detected 1,30,000 cases so far. Roughly 4,000 deaths have been reported. This implies a few things:
– We are picking up only 2.5 to 3 per cent of cases. An underreporting factor of 40, even with relatively high rates of testing in Delhi compared to the rest of India is both a good and bad thing.
– Even if one out of five deaths were reported in Delhi, this gives us an infection fatality rate of roughly 0.4 per cent. (Compare seasonal flu at 0.1 per cent) — that’s a good thing.
– Very high level of asymptomaticity, that is, many people went through the infection without knowing, is again both a good and bad thing. Good, because they did not suffer, bad, because they increased the spread.
- More people having the antibody (Ab) implies that the chance of an active case meeting someone susceptible becomes less. Another good thing.
But I’d like to strike a cautionary note here. One, Ab test positive only indicates past infections. The jury is still out on whether it implies immunity, and, if yes, for how long. My second point of caution would be that if someone is Ab positive, they should still follow social distancing guidelines as we do not know much about their transmission probability and future disease severity, or potential chance for re-infection. When in doubt, protect yourself and others.
What we still need to know about the sero survey is the response and positivity rate stratified by age, sex, job type, district; the sampling design and the sensitivity, specificity of the customised assay used for the test. Just releasing one magic number without a complete report is dissatisfying and is not the best practice for science and policy.
Do regular sero surveys help?
Seroconversions happen and things change over time. Longitudinal snapshots through properly designed seroprevalence studies can be informative particularly when testing capacity is still so limited. The NCDC survey indicates underreporting by a factor of 40. However extensions of the SEIR models can also predict the number of asymptomatic cases to a reasonable margin of error.
We have had projections from government sources, based on the serological survey, indicating herd immunity in metros like Delhi. What do you have to say about that?
Based on the value of the basic reproduction number R0 ranging from 2.0 to 3.5 across the world, herd immunity threshold has been projected from roughly 50 to 70 per cent. This may be achieved over a long period of time, but I would not recommend actively pursuing it in the absence of a vaccine. The sero-survey results are not a license for being cavalier about social distancing and other non-pharmaceutical interventions. Those who are saying we are close to herd immunity in India, please note that India is 1.34 billion people and while the curve is tapering off in Delhi, Mumbai, Tamil Nadu, we have other states on an upward trend. Villages are at risk with almost no healthcare infrastructure. So let us buckle up and do our best to be vigilant, socially responsible and cautiously optimistic.
Even with 0.4 per cent infection fatality rate, if 10 million people in Delhi get the infection there will be 40,000 deaths. Every death count is a life loved by others and a life lost. Let us do our best to persevere till scientists get to a vaccine and treatments for COVID with inspiring research that is moving at a war-time speed.
Several forecasts have predicted that cases in India will peak around August but given how the caseloads are distributed across a few states (a recent health ministry report estimates 86 per cent of the cases come from 10 states), is it possible to predict a national peak?
We wrote a paper in May where we pointed out that the concept of a national peak for a large populous country like India is nebulous. Some states have seen their peaks (like Delhi, the R0 has been below one for three weeks now), and some are coming to boil. There will be a cascade of small and medium peaks across Indian states over the next couple of months.
One of the things that is being said is how countries governed by women have managed to handle the pandemic better. What do you have to say about that?
As the first woman chair of one of the world’s oldest (celebrating 70 years), largest and finest biostatistics departments, I was intrigued by this question. Together with my students, Soumik and Max, we did a world-wide analysis that is getting published this week. It showed that countries with women heads of state were ahead in terms of median performance, by any public health metric you can think of for controlling the pandemic. I think women leaders often have a governance style that is a combination of empathy, efficiency and elasticity, leading to inclusive decision making. This approach leads to effective coordination of different sectors of the government: public health, healthcare, social support and financial sector, rallying together around one common good. However, it is not just the leaders. Societies that are progressive enough to recruit diverse leaders are perhaps doing better in collectively managing the pandemic? Something to think about as we elect our next set of leaders.
Source: Read Full Article