Role of Serological Surveys During a Pandemic: Preliminary Outcomes from the Mumbai Survey

Ullas Kolthur and Sandeep Juneja, TIFR
What is a serological survey and why is it done?
We are constantly exposed to several pathogens and in many cases these do not cause symptoms. Unlike the reported cases from people who have symptoms, serological surveys reveal the proportion of the population that has been infected by a pathogen, with/without symptoms. In sero-surveys scientists typically measure antibodies produced by our bodies in response to an infection. Results of such studies depend on specificity and sensitivity of the assay used to detect antibodies in our blood. 

Why is a serological survey useful for COVID-19 infection?
Sero-survey for COVID-19 infection is extremely useful to know how many people have been infected, with or without symptoms, as well as the geographical spread. Along with other measures, sero-surveys help epidemiologists to learn about herd-immunity. In addition to scientific/clinical value, such studies are extremely useful for governments to adopt policies from health, to opening of schools to commerce. It helps the government allocate epidemic control efforts to where they may be most needed.

Given that mega cities have complex demography and high population density, it is important to assess what aspects of human population and/or their behavior affect exposure to COVID-19 infection. Over the past few months, sero-surveys were done all around the world, focused on the mega cities. Prevalence in New York City in late April was estimated to be 23%, while in London it was about 21% by the last week of May, indicating that about 23% of the population in New York city and 21% in London had Covid-19 infection. In both the cities the disease spread appeared to be winding down at these times. Recently, in Delhi about 23% of the population sampled was found to have been infected in early July. Preliminary results from three wards of Mumbai (Matunga, Chembur and Dahisar) show around 51-60% prevalence in slums in the three wards and 11-18% prevalence in the non-slums in these wards. 

The sero-survey in three wards of Mumbai, performed by NITI/TIFR/BMC, was designed to measure prevalence in (a) general population in slums and non-slums, and (b) health care workers. In addition the study was also designed to reveal age/gender based differences and assess risk factors (co-morbidities, crowding, travel history, etc). Preliminary results, based on 6935 samples, indicate 57% in slums and 15% in non-slum population had COVID-19 infection in the three wards of Mumbai. The population density in the slums can be substantially higher than in non-slums. For instance, the largest slum in Mumbai, Dharavi has an estimated population density of 277,000 persons per square kilometre (Wikipedia), while as per 2011 census the density of Greater Mumbai is about 20,500 persons per square kilometre.  This indicates that crowding and socio-economic factors are potential contributors to the spread of infection.
 Assuming (rather boldly!) that Mumbai slums have 57% prevalence and non-slums have 15% prevalence, and using the census 2011 data for overall slum population in Mumbai (52.5%), one arrives at a ball-park estimate of 37% prevalence for all of Mumbai. This suggests that Mumbai has a higher prevalence of COVID-19 compared to other mega-cities, which could be because of the large proportion of the population residing in dense slums in Mumbai.

 Fatality rate based on the number of reported cases is 5-7%, while the fatality rate based on the sero-survey is around 0.05-0.15%. This difference is because a sero-survey identifies a larger number of infected population than the reported infections; only those with pronounced symptoms seem to be undergoing the RT-PCR based screening and are reported. This suggests that a large proportion of the infected population is asymptomatic or has mild symptoms. The study also shows lower fatality rates in slums compared to the non-slum areas. This could be due to many reasons including  (i) the slum population is younger than the non-slum and they may have better immunity, (ii) the population in the non-slum areas may have a higher prevalence of co-morbidity. These hypotheses need to be tested and have immense value in fighting the health crisis. This serological survey involved the detection of one of the antibodies against the COVID-19 infection and currently it is not clear how long after the infection can one detect the antibodies. Hence, the sero-survey needs to be repeated on the same set of people and also on a larger population. 

Mumbai sero survey -
London sero survey -
New York sero survey -

[Last Updated 18 August 2020]