Covid-19: India Must Urgently Manage Regional Contact Networks
People throng to a market during a nationwide lockdown, imposed in the wake of coronavirus pandemic, in Mumbai on March 28, 2020. (Photograph: PTI)

Covid-19: India Must Urgently Manage Regional Contact Networks


The impact of Covid-19 has been felt far and wide across countries and continents. With the virus gathering rapid spread and rates of cases and mortality exploding multi-fold each passing day especially in advanced economies such as the United States, Italy and Spain, India is not far behind and currently stands at 4,067 cases as of April 4. This has significant multi-fold ramifications on the cultural, economic and health infrastructures within and across different states over the next few weeks and months.

With several mitigation procedures in place such as lockdown, and travel bans, the spread will potentially be slowed and flatten the—now ubiquitous—curve to allow health care providers to provide effective care to those in critical need. The critical question remains what the future holds, both in the short-term, say 2-3 months, and the long-term. Using rapidly evolving data emerging from India, we attempt to answer these questions in a rigorous evidence-based manner.

At a fundamental level, epidemics spread over a network – where carriers spread diseases with whom they are connected to some degree. These networks or 'links' can be via person to person interaction, local transport or travel, or via shared spaces/ surfaces. The geographic location of the networks is relevant because there may be different modes of spread that exist within any region.

The mitigation procedures effectively break this human to human 'link' to disconnect the communities, and hence reduce the spread. Classical epidemiological models attempt to estimate a reproduction number or R0 (pronounced as R-naught) to describe the intensity of the outbreak and is defined as the average number of infectious individuals generated by one infected individual in a fully susceptible population.

Emerging studies have put the R0 number for Covid-19 for India to be around 2 – essentially doubling per each infected individual.

With advanced modeling techniques, we have built a network-based forecaster for predicting the number of infected cases over the next three months over specific geographical contact networks for every state in India. Using reported data at an individual level we first created a national-level network across India of the patients infected with Covid-19. In this network, the patients are the nodes and the two patients are connected if they are proximate i.e. in the same geographical location, by latitude and longitude.

We then extract the state-level network—determined by the location of detected nodes—and that becomes the initial infected network of the state. We then ‘complete’ this initial network with the number of individuals who are ‘susceptible’ to the disease using the population density of the state i.e. likely to get infected. In essence, higher density implies more chances of interactions—social or physical—leading to a faster spread of the infection. We estimate the rate of infection in these susceptible individuals using a network-based ‘susceptible-infected-removed’ model that uses the reproduction number to forecast the number of affected individuals in the future.

The estimated curve closely follows an exponential growth under a reasonable estimate of the reproduction number (R0 = 2). The first chart shows this estimated number (on the vertical axis) as a function of time over the next three months (April-June, on the horizontal axis). The second chart is the actual projected number of individuals who are likely to be infected at two cut-points in this curve – May 15 and June 1, 2020, for each state.

<i>(This is based on network </i><i>‘</i><i>susceptible-infected-removed’ models. The initial network for each state contains the infected individuals, and an additional number of susceptible individuals so that the total size of the network matches the population of the state. Edges are defined by whether two persons are located within a band of three degrees of latitude and longitude of one another. For running the simulations, the reproduction number R0 was taken to be 2, and the average time at state I was taken to be 10.)</i>
(This is based on network susceptible-infected-removed’ models. The initial network for each state contains the infected individuals, and an additional number of susceptible individuals so that the total size of the network matches the population of the state. Edges are defined by whether two persons are located within a band of three degrees of latitude and longitude of one another. For running the simulations, the reproduction number R0 was taken to be 2, and the average time at state I was taken to be 10.)

Some startling numbers emerge. Most states will see a significant rise in the number of cases starting mid-May and the top hotspots are Maharashtra, Kerala, Karnataka, Telangana, Uttar Pradesh, Rajasthan and Gujarat with Delhi, Andhra Pradesh, Tamil Nadu, and Madhya Pradesh close behind. While initial hotspots could be traced to international travel and port of entries, some newly-formed hotspots have come up which can be potentially attributed to hub-and-child nodes of contact networks in large public gatherings.

While these predicted numbers are in the hundreds over the next few weeks, they rise to thousands and then to lakhs by June – if the current rates continue.

The chart below shows the number of affected cases for May 15 and June 1, where some states will cross a lakh; and this under a conservative estimate under a reasonable adherence to the lockdown. Unsurprisingly, states with high population density and with ports of entry are the ones that will see a surge and perhaps would need to be prepared the most.

<i>(Same network SIR model as above)</i>
(Same network SIR model as above)

What Measures Can Be Taken

States, especially the hotspots, should consider implementing state-level policies that are in keeping with their governance structure, culture, and health system.

This should include:

  1. Extensive screening and contact tracing deployed at a community level,
  2. Strict social distancing and self-quarantine policies,
  3. Enforced quarantine policies leveraging community resources like hotels to isolate asymptomatic positive cases.
  4. A rapid ramp-up in healthcare capacity rapidly in both first-tier and second-tier cities by de-canting non-critical services and extending critical care capabilities in existing hospitals, building temporary facilities for rapid treatment, and leveraging alternate care facilities (like hotels, convention centers, etc.).
  5. State-specific identification of disease hot-spots, rapid contact tracing and testing, and sufficient medical supplies and protection for the health workforce.
If we want to reduce the cumulative effect of the contact network, then we have to reduce contact, and this includes the need for an extensive commute to get adequate healthcare.

Impact On The Economy

The Indian economy was already going through a rough patch before the arrival of the pandemic. India went into a nationwide lockdown for 21 days, starting March 25, 2020, that was announced the previous evening. With this nationwide lockdown in force along with state or region-specific curfews in place over and above the lockdown, the tremors are currently felt almost everywhere. A pre-lockdown survey report released by FICCI on ‘Impact of Coronavirus on Indian Businesses’ presented a grim picture of the business scenario. It had reported a significant reduction in orders and cash flow, a hit in the supply chain, and an overall downslide effect on businesses. The travel and hospitality sector has suffered one of the harshest blows, and airline companies are on the verge of facing bankruptcy, with global and domestic travel networks being shut.

While the pandemic itself has its own way of taking a toll on the socio-economic health of a nation, the counter-measure of stopping the spread via a total lockdown of the country, if extended for months, is something which will have harsh ramifications. With factories being shut, several places in the country are reporting non-availability of workers engaged in manufacturing essential commodities including medical equipment. A considerable chunk of India’s workforce comprises the unorganised sector and migrant workers, those who had left their peri-urban or rural base to work as daily wage-earners in urban areas, often in other states. The lockdown has thrown them out of work in a short notice.

Furthermore, the idea of social distancing via lockdown already got squashed, when a deluge of these workers tried to return, packed in trains, buses, and in the absence of them, on foot — walking hundreds of miles to their homes, with the uncertainty of feeding their families looming large. A considerable chunk thus became susceptible to the infection with a possible addition to the contact network in their localities.

The central government and several state governments have already taken several steps to mitigate this crisis, including economic packages to support the poor and financial stimulus, but there are several criticisms arguing that the aid is lower than 1 percent of the country’s GDP. The Ministry of Health and Family Welfare in a recent advisory has outlined several steps to tackle this crisis, which includes setting up quarantine facilities for the migrant workers and regular surveillance of their health, among others.

A lockdown is arguably a strategy to buy time to check the pressure on the healthcare system, but the number of infected individuals is expected to surge once the testing rate is significantly increased given that there are potentially many cases that are undetected.

State-specific identification of disease hot-spots and strengthening the healthcare system, providing sufficient medical kits to the health workforce are what we need to look forward to in the coming months.

A countrywide lockdown is not a long-term viable solution, and a trade-off between life and livelihood has to be adopted very carefully in this context.

India may have some cultural, environmental and inherent genetic factors that give it an advantage – deeper understanding of these, while being hyper-prepared for different scenarios, will allow India to emerge as a leader in this global fight against the coronavirus. The rate of spread of Covid-19 in India is still in its early phase and we believe proactive measures, as opposed to, reactive measures are the need of the hour – for the administration and general population to deal with this pandemic, and others in future.

Veera Baladandayuthapani is Professor of Biostatistics, Rupam Bhattacharyya is a doctoral student in Biostatistics and Shariq Mohammed is a postdoctoral research fellow, at University of Michigan. Sayantan Banerjee is Assistant Professor, Operations Management & Quantitative Techniques Area, at IIM Indore. Upali Nanda is a Principal and Director of Research at HKS and an Associate Professor of Practice in the Department of Architecture at University of Michigan.

The views expressed here are those of the authors and do not necessarily represent the views of BloombergQuint or its editorial team.

BQ Install

Bloomberg Quint

Add BloombergQuint App to Home screen.