Modeling the Impact of Lockdown Measures and Coronavirus Response in India

people on street in India

Q&A with Rupam Bhattacharyya

PhD student in Biostatistics

April 8, 2020

Click Here for the Latest on COVID-19 from Michigan Public Health Experts

Rupam Bhattacharyya, a doctoral student at Michigan Public Health, is part of a team of researchers that, as the coronavirus pandemic unfolded around the world, used standard epidemiologic models to do a situational assessment of the crisis in India—providing real-time data for authorities to assist leadership in addressing this global pandemic.

Based on the report, the team published an analysis of their data highlighting the benefits of social distancing. Now the team has released a second article and made an online dashboard available to the public that utilizes emerging data from India to produce updated projections. 

Bhattacharyya discusses the team's work, his contributions, and his own interests in public health.

How did you become involved in this study?

When the coronavirus pandemic started unfolding, I became really interested in getting involved in work that could help address it. I was able to join this study through my advisor, Veera Baladandayuthapani, and the chair of the Biostatistics department, Bhramar Mukherjee, who is also the principal investigator on the study. Though for my own research I usually work in different areas—precision medicine and cancer—I have always been interested in epidemiology and this was an opportunity to shift my skills toward virology and spread. It is a new avenue for me and I am really enjoying it.

The study, itself, is specifically interesting to me in two ways. First of all, the team and data are focusing on India, which is where I am from, originally. Second, as I watched the data come in and saw that this was truly a public health crisis, I realized that this was a situation that I and other data analysts could contribute to immediately. In this epidemic we can help design studies that can focus the situation, help others learn and respond to it, and aid in the eventual stopping of the virus. I feel very fortunate to be a part of that.

How did the study come about?

Before we started the study, we saw that there was a lot of modeling work being done on the virus and its spread in other parts of the world, specifically China because that is the epicenter for the virus. But we felt that similar attention was needed for other parts of the world. India is particularly important to consider because it has a very high population with a very high population density. One sixth of the world’s population lives in this very condensed region, so if the pandemic really grows there, it will mean trouble for not only the country or the Asian continent, but the whole world. By doing modeling before the pandemic grew out of control, we hoped that we could provide a message and information about what the future might be like and the kinds of actions that needed to be taken to curb the pandemic.

What does the study hope to accomplish and what is your role in it?

After we had analyzed responses to similar outbreaks in India, and then reviewed responses to the coronavirus pandemic in other countries, we were interested to determine the outcomes that could arise in India—what the spread would look like—if the country undertook various levels of intervention, from the most mild to the most extreme. This is the portion of the study that I have been most involved in. Using an existing mathematical paradigm and the data we could gather when this part of the study began in mid-March, the study forecasts the future infection numbers and hypothetical scenarios in India for interventions that span the spectrum of intensity—what happens, for example, if the country doesn’t do anything to curb the spread, what happens if it does something mild like a travel ban, and what happens if it employs strategies that are very extreme? 

To help accomplish this, I worked on the kinds of parameters we were feeding into our mathematical models, how we could use the data we had in more effective ways, and the various ways we could interpret the numbers we were producing. The numbers themselves are very dynamic and change every day, so it was—and continues to be—important to pay close attention to emerging patterns.

What are the takeaways from the study so far and what’s next?

Overall, the data has shown us that it’s very obvious that more and more extreme interventions can help the spread slow down, but that it is also very important to determine exactly when and where it’s prudent to take these extreme measures. Nothing happens in a vacuum and the kinds of measures we take to combat the pandemic can have wide-ranging consequences for a country and its people. We see that play out in the negotiation between public health and economics—something like a prolonged lockdown can actually destroy some peoples’ livelihoods, but without a lockdown, how many lives are put at risk?

After we concluded the forecasting, we added an economic piece to our work with contributions from experts in India and across the globe, with the hope that the result could address this balance between public health and economic concerns. Now, as new data comes in from India that shows the effect of the lockdown and other interventions, we have also begun to update our predictions and forecasts so we can understand the impact of the lockdown. That information has been published on Medium. 

We have also set up a web dashboard, available to everyone, that incorporates emerging data each day from India into our forecasts. We hope that this will be useful to a lot of people, especially from a policy perspective, as officials try to determine where they should be allocating resources and what kinds of interventions need to happen in the future.

When you’re not working on this study, what areas of public health are you most passionate about?

I come from a very theoretical background, mostly mathematics and statistics. In my public health research before I arrived at Michigan, my focus was on biology because I had always been interested in things like genetics. But at Michigan, I became deeply interested in precision medicine. It initially came as a new thing to me. I didn’t have a good idea how a statistician could be of help in that arena. But the more I got into it, the more I could see how my skills could be of use. Now I work primarily in oncology and precision medicine, and I’m focused on how we can optimize cancer treatments for specific individuals. Cancer is a very personal disease and no two tumors are the same. Even if two patients have the same kind of tumor in the same organ, the tumors are inherently really different from a cellular perspective. I build Bayesian machine-learning models that take sets of patients and their genomic characteristics and help predict how they’ll react and respond to certain medications. Armed with that data, physicians can decide on drugs and drug combinations that will ensure the best outcome for a patient.

Tags