The ongoing spread of the coronavirus disease 2019 (COVID-19) that originated in Wuhan, China, has caused more than 400,000 confirmed cases and more than 19,000 deaths across 172 countries. The disease has death rates as high as 10-27 percent in people over 85, and 3-11 percent among people aged 65-84 years. In Bangladesh, there have been 44 officially confirmed cases of COVID-19 as of March 26, but the true burden of the disease is likely to be significantly higher. So far, only a few hundred tests have been reportedly administered in the country. Without widespread testing, the true burden of the disease and the extent of community transmission cannot be determined.
With the sharp increase in cases globally, there have been urgent calls in many countries for travel bans and "social distancing" practices. In Bangladesh, all educational institutions, and private and public offices have been temporarily closed in an attempt to halt the outbreak. Travelers from abroad have been asked to self-quarantine, and all domestic flights and public transportation have been suspended. Despite these measures, considerable volumes of travel are occurring daily within Dhaka, and from Dhaka to the rest of the country, with a spike in travel following the announcement of the lockdown.
The virus that causes Covid-19 is spread by close contact. Experiences in South Korea, Singapore, and Hong Kong have shown that widespread testing coupled with strong social distancing measures that limit close contacts, such as school closures and banning of public gatherings, can be effective in controlling the spread of the virus. Indeed, these are some of the few interventions available to us to slow the exponential increase of new cases by reducing transmission, if these are deployed early. During the 1918 influenza pandemic, for example, cities that intervened early had significantly lower death rates.
Social distancing measures are particularly important in the context of large urban hubs and megacities. Dhaka is one of the densest and fastest growing cities in the world, with more than 18 million inhabitants in the greater Dhaka area. It serves as a central hub with high connectivity to the rest of the country and internationally. Understanding what areas within Dhaka city might serve as transmission hotspots and how to implement strong social distancing measures is, thus, critical.
Population movement and aggregation have been shown to be important drivers of epidemics. However, day-to-day population movement patterns are difficult to measure using traditional sources of population data. Anonymised mobile phone call detail records (CDR) provide a unique opportunity to estimate population dynamics in real time at a high spatial and temporal resolution. We analysed population movement from anonymised mobile phone data, from Telenor Group's mobile operator Grameenphone in Bangladesh, for about 60 percent of the population in Dhaka city to identify potential "hotspots" for disease transmission. This builds on our previous work with Grameenphone and the National Malaria Elimination Programme, where we used mobility estimates derived from CDR data to map the spread of malaria (eLife, 2019).
Our analysis shows that in certain areas of Dhaka city, the average daily number of people traveling into that location can exceed 30,000. These include Gulshan, Uttara, Shahbag, Motijheel and Paltan areas. When we consider density, places such as Motijheel are particularly important given the larger volumes of traffic into a relatively small area. When we analysed the travel network in Dhaka, we found that certain parts of the city, such as Dhanmondi, Gulshan, Motijheel and New Market, were more highly connected to other locations and have large volumes of incoming traffic from all across Dhaka city. Again, the high connectivity of these locations suggests that they could serve as potential transmission hotspots. During the 2017 Chikungunya outbreak, we found in our research (bioRxiv, 2019) that these areas did in fact have relatively higher disease incidence.
Our analysis also shows that large volumes of daily travel occur between Dhaka and all other regions in the country, particularly on weekends. We estimate that, on average, almost 400,000 people travel daily to other parts of the country. This suggests that an outbreak originating in Dhaka is likely to spread rapidly to all other parts of the country. The rush of people leaving Dhaka following the announcement of the lockdown is, thus, especially concerning, and the disease is likely to spread across Bangladesh. Again, without immediate widespread surveillance, it will be difficult to estimate the spatial spread of Covid-19 in Bangladesh.
The time to act is now. We can confidently say, from the experiences of other countries, that acting decisively and quickly is the best way to prevent the health system from being overburdened. Bangladesh health authorities should immediately expand their surveillance capacity so they know the extent of transmission of the virus, and they should enact social distancing measures, such as banning large gatherings, reducing unnecessary travel, and staying at home as much as possible—all of which have been shown to be effective in slowing the epidemic in other countries.
Reducing travel overall and especially within high-density, highly-connected locations around Dhaka and the related shutdowns will be critical for slowing the spread of the epidemic. If the infection is allowed to spread unchecked, the health system of Bangladesh is likely to be overwhelmed with severe cases needing hospitalisation, as we have seen in Wuhan, China, and across Italy, with devastating social and economic impacts.
Dr Ayesha Mahmud is an Assistant Professor of Demography at the University of California, Berkeley. Dr Caroline Buckee is an Associate Professor of Epidemiology at the Harvard T. H. Chan School of Public Health.