In May, India potentially had 6.4 million COVID-19 infections and for every confirmed case there were 82-130 infections that went undetected, according to the results of the first sero-survey conducted by over 50 scientists of the Indian Council of Medical Research (ICMR) and other institutions.
These findings that now formally appear as a peer-reviewed paper, describe a nearly month-long investigation to check the spread of infection when India was in the midst of a lockdown, were known to the government at least by early June. However, details — such as the probable all-India infections — have only now been made public after the paper was published late on Thursday in the Indian Journal of Medical Research, the Council’s in-house medical journal. As of Friday, India has officially confirmed 4.5 million cases and 76,000 deaths.
Also read: Banking on serology: The Hindu Editorial on seroprevalence studies
Serology surveys involve testing blood samples for the presence of antibodies, that generally form within 2-3 weeks of an infection. These tests are faster than the gold standard RT-PCR tests but usually cannot detect an active infection, which the latter can.
At a press conference on June 12, ICMR Director General, Balram Bhargava had said , “0.73% of the population surveyed in 83 districts was likely infected by the virus” but he didn't specify what the all-India prevalence was. The survey had identified 83 districts but eventually sampled participants in only 70 districts.
Director of the National Institute of Epidemiology Manoj Murhekar, who is the lead author of the study, told The Hindu that the proportion today of those confirmed and those who've probably been exposed to the infection would be still be high but significantly reduced from May.
“The serosurvey conducted recently in Chennai shows that there were about 25 infections for every RT-PCR confirmed case (or 1:25) and this is a city with a high degree of testing. Testing was much lower in May. I'd say the ratio now would certainly be below 1:80.”
Also read: Meta-analysis does not support continued use of point-of-care serological tests for COVID-19
The ICMR is in the midst of another serology survey in the same villages and cities, the results of which are likely by the end of the month. “The prevalence of the infection will naturally be much wider,” said Dr. Murhekar.
For the survey from May 11-June 4, 28,000 adults from 21 States across 70 districts were tested for the presence of SARS-CoV-2 antibodies. The 290 who tested positive for antibodies were tested with the Covid Kavach Elisa Kit , developed by the ICMR-National Institute of Virology. The scientists used another ELISA kit, called Euroimmun ELISA to recheck the 290 and found only 157 positives. The two kits were used, the authors say, to improve the accuracy of feretting positives.
“The pooled adjusted prevalence of SARS-CoV-2 infection was 0.73% at the national level... amounting to 6.4 million infections in total by early May 2020,” the authors note.
News Analysis | India may have undercounted cases
“The findings of our survey indicated that the overall seroprevalence in India was low, with less than 1% cent of the adult population exposed to SARS-CoV-2 by mid May 2020. The low prevalence observed in most districts indicates that India is in early phase of the epidemic and the majority of the Indian population is still susceptible to SARS-CoV-2 infection,” the paper said.
The latter point was underlined by Dr. Bhargava, too, in June.
The survey also showed that the infections, between May and June, had already spread to rural India. Of the 157 who tested positive, 109 were from villages and 48 were from urban localities, underlying that despite the strict lockdown all of April and May and the high prevalence in cities, the infection had spread quite substantially beyond. Dr. Murhekar explained that this was because nearly 75% of the population sampled were in villages. “The positivity (odds of findings a positive case for every test conducted) was much higher in cities. The survey was also designed to check for the geographical spread of the infection then,” he added.
The 70 districts were categorised into four strata according to the incidence of reported COVID-19 cases per million population (zero, low=0.1-4, medium=5-10, high greater or more than 10) as on April 25, 2020. There were nearly 1,09,000 infections in the 'zero' stratum regions and the maximum — 3,91,941 infections — were in the 'medium' incidence districts. More positives were found among men than women and there were 130 positives younger than 60 and 27 above 60. Those in 'occupations with higher risk of exposure' were 39% more likely to be infected than those who weren't.
The Infection Fatality Rate (IFR) or deaths per estimated infections, ranged from 0.18 in zero stratum and 11.72 in the high stratum districts, for every 10,000 infections.
“Considering that the death reporting in India is incomplete, and differences in access to testing facilities across districts necessary for declaring the COVID-19 confirmed deaths, the present IFR (infection fatality ratio) is likely an underestimate. While the overall IFR based on the serosurvey findings was much lower than that reported from Santa Clara County, U.S. (0.12-0.2%) 16, Iran (0.08-0.12%) 23, Brazil and Spain (1%) 24, the IFR from the high-stratum districts,where reporting is assumed to be more complete, was similar to those reported above,” the authors noted.
The Hindu had reported on June 14 that 7,00,000 people could have been infected by the SARS-CoV-2 virus in these districts even in early May and that the number infected was at least 20 times more than what official confirmed cases indicated.
