New Delhi: The long-awaited findings from the first national seroprevalence survey, which researchers from the Indian Council for Medical Research (ICMR) conducted during May and June this year, were finally published in the Indian Journal of Medical Research on September 10, 2020.
The document repeats what ICMR officials said in their June 11 press conference, when they first released the survey results: that 0.73% of adults in India had been exposed to the new coronavirus, which which translates to 6.4 million infections as of early May. This is in stark contrast to India’s cumulative case load on May 7 of 52,592. However, the document has much more detail that allows a more accurate interpretation of the survey results.
ICMR researchers conducted the survey from May 11 to June 4. During the survey, they allegedly tested blood samples obtained from 28,000 people in just over 80 districts for immunoglobulin G antibodies using the COVID Kavach ELISA kit. Based on the results, they said that they had estimated the national prevalence of new coronavirus infections at that time to be 0.73% of the population and an infection mortality rate (IFR) of 0.08%.
When these details were first released on June 11, they were met with surprise, not least because the authors had decided to release the results before the document containing more detailed analyzes, thus lending the claims to a independent verification, was available even as government officials. began expanding on the results to back up its own claims, especially that the national shutdown, which began on March 24, had been successful. However, as of today, the case growth graph does not show any effect on the slope of the curve due to blocking.
The decision to hold a press conference prior to the publication of the document also contributed to confusion about how ICMR had calculated the IFR, what statistical methods they had used to arrive at the other results, what was the specificity and sensitivity of the kit that ICMR had used, the age breakdown of the results (especially since the survey had excluded people aged 17 and under) and the confidence intervals of different data points of interest.
(As Priyanka Pulla reported for The science of wire in May: “To account for… variations, a diagnostic test should ideally be tested with multiple groups of people. And when this is not possible, the researchers calculate the range through which the sensitivity and specificity values can vary for each kit, in addition to single estimates such as 100% or 98%. This range is called the confidence interval ”or CI. Similarly, in the current context, ICMR researchers did not report on June 11 the minimum and maximum between which the estimated case burden could vary.)
According to the published article, the 28,000 people had enrolled in 30,283 homes the researchers had visited, in 700 villages and districts in 70 districts in 21 states. (Interestingly, ICMR chief Balram Bhargava had said that 83 districts had been surveyed during the press conference.) The survey researchers used the ELISA COVID Kavach kit developed by ICMR to analyze the samples; these kits are 97.9% specific and 92.37% sensitive.
The samples that tested positive were retested with another kit, called the Euroimmun SARS-CoV-2 ELISA, which, according to the company’s press release, is more than 99% specific (specificity is the rate of “true negatives” ). Earlier, Prabhat Jha, an epidemiologist at the University of Toronto, had told him The science of wire that any kit with less than 99% specificity would be “quite useless in most seroprevalence studies.” ICMR had first reported in mid-May that its Kavach kit was 100% specific, a figure which it later revised to 97.9% after questions from The science of wire.
The authors of the article have acknowledged this problem and write:
The estimated seroprevalence is a function of the sensitivity and specificity of the serological tests. Appropriate thresholds for sensitivity and specificity are influenced by the prevalence of infection. As was done in our study, the use of two tests sequentially under the condition of positive result in both tests would lead to an overall increase in specificity at the cost of reducing sensitivity. The sequential use of COVID Kavach and Euroimmun ELISA allowed us to potentially reduce the false positive up to 0.01% obtaining a serial specificity of 99.99% (if the independence between the tests is high).
However, they paid a price for this improvement: the serial sensitivity (that is, the sensitivity, or “ true positives ” rate, corresponding to a sample tested in both kits, one after the other) was reduced to 86, 67%, below the sensitivity of using only the COVID Kavach kit, 92.37%.
Let’s say 10,000 people are infected. Kavach’s kit would correctly identify 9,237, but using both kits together would identify only 8,667 and classify the remaining 1,333 as “negative”. In an ideal world, the number of “negatives” should be zero, since 10,000 are infected. So now, 1,333 cases have been “lost”, at the expense of the survey that excludes, as is desirable, uninfected people to the tune of 99.99%.
As the authors of the article write: “More specific testing is preferred in a low-prevalence setting like ours to minimize the large number of false positives.” That is, ‘it’s okay to overlook some positive cases as long as we leave out as many negative cases as possible’.
Of the 28,000, 290 tested positive; when they were retested, the number dropped to 157. The document indicates that, of these, 109 were from villages, 23 were from urban areas that were not slums, and 25 were from urban slums. Given the unequal urban-rural division of the participants, 0.52% of those sampled in rural areas and 0.66% of those sampled in urban areas tested positive. This indicates, among other things, that India’s COVID-19 epidemic had spread to rural areas, despite multiple claims at the time that it was still confined to urban foci.
However, as epidemiologist Chandrakant Lahariya has warned, extrapolating the survey results to all of India may not be that straightforward as most of India’s cases are currently in urban centers, while 73.8% of the people in the survey were from rural areas.
In the final analysis, the researchers write:
6,468,388 cumulative adult infections (95% CI 3,829,029-11,199,423) were estimated in India at the beginning of May. The general [infection to case ratio] was between 81.6 (95% CI: 48.3-141.4) and 130.1 (95% CI: 77.0-225.2) with May 11 and May 3, 2020, as plausible benchmarks for reported cases. The IFR in the surveyed districts of the upper stratum, where the reporting of deaths was more robust, was 11.72 (95% CI 7.21 to 19.19) to 15.04 (9.26 to 24.62 ) per 10,000 adults, using May 24 and June 1, 2020 as plausible. benchmarks for reported deaths.
The infection ratio per case implies that for every person who tested positive for an RT-PCR test and entered India’s official case count, we missed 81.6 to 130.1 people who also tested positive for the survey period.
Furthermore, seropositivity, that is, the fraction of individuals in a group who tested positive, was found to be the highest among adults aged 18 to 45 years (43.3% among those who tested positive, 0, 50% among the individuals of this age group in the survey), it is followed by those between 46 and 60 years (39.5% and 0.65%) and the lowest among those over 60 years (17.2% and 0.55%).
“Our survey findings indicated that overall seroprevalence in India was low, with less than one percent of the adult population exposed to SARS-CoV-2, as of mid-May 2020,” the researchers wrote in the final section. Given the period in which the survey was conducted and the incubation period of the virus, its results correspond to people who were infected around the end of April. “The low prevalence observed in most districts indicates that India is in an early phase of the epidemic and the majority of the Indian population is still susceptible to a SARS-CoV-2 infection.”
According to PTI, the document also highlighted the need to continue to implement context-specific containment measures, including testing of all symptomatic patients, isolating those who test positive, and monitoring high-risk contacts to slow progression of the disease. epidemic and prevent the growing burden of cases from being overwhelming. the health system.
At the June press conference, ICMR’s Bhargava had mentioned that the survey had classified the 70 districts from which people had been surveyed into four ‘strata’, based on the number of confirmed cases there as of April 25. While it did not provide a stratum, on a smart breakdown the document is clearer: 15 districts were classified as ‘zero cases’, 22 districts as ‘low’, 16 districts as ‘medium’ and 17 districts as ‘high’. The seroprevalence in the four strata ranged between 0.62% and 1.03%.
However, the researchers cautioned that four of the 15 districts in the survey (and 233 of those districts in India, according to the paper) with ‘zero cases’ did not have COVID-19 testing labs at the district headquarters and that samples collected in the area for testing had to be transported to the state headquarters hospitals. “Current findings of seropositivity in strata of districts with zero to low incidence of COVID-19 cases underscore the need to strengthen surveillance and increase testing of suspected cases in these areas,” the researchers conclude.
On July 10, PTI had reported that ICMR would conduct “a nationwide seroprevalence survey to determine population exposure to the new coronavirus… as a follow-up” to the one conducted in May-June.
With PTI inputs
