India has launched a repeat nationwide survey to measure changes in the new coronavirus infection’s prevalence after a three-month gap, and probe for the first time prevalence among adolescents to guide school-opening policies.

The survey by the Indian Council of Medical Research, the apex health research agency, will look for antibodies to the coronavirus in 28,000 blood samples — 400 each from the same 70 districts where the council had conducted a similar survey in May.

“We’ve embarked on the second survey — this is a work in progress. It should be completed by the first week of September,” said the council’s director-general Balram Bhargava.

But the repeat survey will sample all members of households, including children and adolescents aged 10 years or above.

The first survey had sampled one member from each household and sampled only people aged 18 years or older.

“Prevalence rates among children or adolescents could have implications for school-opening policies,” a senior clinician-researcher who helped design the survey told The Telegraph. “It is not the only factor, but it could help guide decisions.”

High prevalence among children and adolescents would support school opening, while low prevalence would imply greater risk, the researcher said, but asserted that the prevalence rate would not be the only criteria to determine school opening.

In Bengal, the survey will screen households in Calcutta, Alipurduar, Bankura, Jhargram, East Midnapore and South 24-Paragnas.

The ICMR has not yet made public the details of the first survey, although it had on June 11 revealed that the average prevalence across the 70 districts was 0.73 per cent. But researchers familiar with the findings have said some of the districts sampled had up to 25 per cent prevalence.

Bhargava said on Tuesday that the first survey’s findings had undergone two rounds of peer review and were expected to be published in the Indian Journal of Medical Research this week.

The repeat survey is expected to provide insights into the rate of spread of the infection in specific districts amid evidence from select cities that millions of people nationwide have been infected and have recovered without knowing it.

Health researchers point to the increase in the prevalence in Delhi within a month.

A survey by the National Centre for Disease Control in Delhi had in mid-June found that 23 per cent of the population had antibodies to the virus. But an independent survey by the Delhi state government in mid-July found antibodies in 29 per cent of the sampled population.

The 6 per cent increase in infection prevalence applied to Delhi’s estimated population of 20 million would imply that hundreds of thousands of more people had become infected unknowingly during the period between the two surveys.

While health authorities are tracking daily counts of patients with Covid-19, 80 per cent of infected persons are expected to have only mild symptoms or no symptoms. Such antibody surveys are intended to provide the true burden of infections.

India on Tuesday recorded 60,975 new Covid-19 cases, raising the total number of infections to over 3.16 million, of whom more than 2.4 million have recovered, 704,348 patients are under medical supervision and 58,390 have died.