Delhi’s Sero Survey Data Suggests Covid Cases And Deaths Much Higher Than Govt Figures

The BJP has been alleging that the Delhi government has been hiding the actual number of Coronavirus cases and deaths. Now the Sero survey, that noted that over 23% of Delhi’s population has been exposed to Covid-19, indicates there is a contradiction between Government’s data and the actual numbers.

Jeevan Prakash Sharma

Updated on: 23 July 2020 7:25 pm

At least three BJP-run municipal corporations in Delhi had raised alarm bells in early June regarding mismatch in data of Coronavirus-related deaths, released by AAP government, and the bodies of Covid patients at crematoriums.The municipal bodies alleged that the number of deaths was double the government data.

A Sero surveillance study conducted by the National Center for Disease Control (NCDC), in collaboration with the Delhi government, suggests that total Covid cases in Delhi are much higher than what the official records show.

The survey, conducted on 21,387 people from 11 districts, shows that 23.48 per cent people had already contracted the virus by July 10. However, Delhi government data shows that total positive cases by July 10 were 1,09,140. If the approximate population of Delhi is 3 crore, according to a World Health Organisation report, then just 0.36 per cent of the total population was infected by July 10, according to the government data.

Comparing the survey data to the actual number of Covid cases, one can conclude that the infection was 65.22 times more (23.48/0.36) than what the government recorded.

But can this application be extrapolated to the number of Covid deaths in Delhi?

According to government records, cumulative deaths until July 10 were 3,300. If the Sero survey data is extrapolated to Covid deaths, which showed that the actual number of infections was 65.22 times more than the government data, then the total deaths due to Covid in Delhi would be 2,15,226.

Let’s look at it from the government’s own fatality rate. The 23.48 per cent (Virus spread) of the total population of Delhi translates into 70 lakh. Going by the Delhi government’s 3% fatality rate claim, the total deaths in Delhi would be 2 lakh 10 thousand. This is much higher than the annual death numbers. According to the annual report on registration of births and deaths in N.C.T. in 2018, 1,45,533 people died in Delhi in 2018.

So does it mean that the Sero surveillance data about the spread of infection is also incorrect? Or did over 2 lakh people actually die in Delhi due to Covid-19 till July 10? Can there be a direct correlation between the number of asymptomatic or unrecorded cases and the number of deaths?

Data-driven policy making firm Sapio Analytics Co-founder and CEO, Ashwin Srivastava, feels that the sample size and distribution of the Sero survey is not good enough to come to a conclusion for the entire city and a direct extrapolation must be avoided, but it does reflect a much higher penetration of the virus.

“Our evaluations backed by this new-found context show at least a 100% higher number of deaths than what is officially recorded,” Srivastava said.

According to him, “This indicates what we have always believed in and shared for planning purposes, using data-driven algorithms, that the actual number of cases with exposure to COVID-19 virus is significantly higher than the recorded numbers.”

“Higher number of exposed cases than the recorded cases was seen in Wuhan as well, where a very strong real-time testing was done. So, it's nothing to panic about,” he adds.

Eminent epidemiologist and health economist, Dr Ramanan Laxminarayan, says that the 3% fatality rate “doesn't make any sense”.

“Cases and deaths are both being undercounted because of low testing levels. Even if infection fatality rates were a modest 0.3%, that would indicate four times the numbers of Covid deaths as having been reported. Given that a patient would have had to have a Covid test in order to be certified as a Covid death, it is likely that the number of deaths attributable to Covid have been undercounted,” Laxminarayan said.

Dr Rahul Bhargava, Director, Department of Clinical Hematology & Bone Marrow Transplant, Fortis Hospital, Gurugram says that there are two possibilities only: either the Sero surveillance data is not right or the virus is mild in India.

“As such, in infectious diseases of such nature, calculating fatality rate will be flawed as it will depend on enrollment which can vary from country to country -- their guidelines and methods of testing. There is no doubt that this disease in India has not been that severe or fatal till now, which is a good thing,” Dr NN Mathur, Director, Lady Hardinge Medical College (LHMC), said.