OFFICIALS of the Union Ministry of Health and Family Welfare (MoHFW) suffer from foot-in-mouth disease every time they brief the media on COVID-19. If they do not wish to be transparent about the entire data and prefer to make facts public only selectively, that becomes clear more often than not. This time it showed in the June 11 briefing by the Director-General of the Indian Council of Medical Research (ICMR), Balram Bhargava, on the nationwide sero-surveillance survey the ICMR conducted in association with the Chennai-based National Institute of Epidemiology (NIE) and National Institute for Research in Tuberculosis (NIRT) and in coordination with the State health departments and other stakeholders.
As reported earlier by Frontline (“Denial and Deception’’, June 19), the ICMR had announced on May 12 that in order to get an estimate of the prevalence of infection in the general population it would conduct community-level IgG antibody-based serological testing in 69 districts randomly selected from 21 States (Fig. 1). This was supposed to cover 24,000 adults distributed across four strata of the districts categorised on the basis of reported cases of COVID-19. The survey involved collection of 3-5 ml of venous blood samples from 400 randomly selected individuals (one per household) from 10 clusters in each district.
IgG tests give you data about the prevalence of infection up to about a fortnight earlier since IgG antibodies begin to appear in the infected host only after 10-14 days and are detectable for even up to a few months after the infection has passed in the host. Assuming that the survey would have begun around mid May, the sero-surveillance data would tell you the state of prevalence of infections around April 30 and a few days after. If the blood sample collected is positive for IgG antibodies, the person had SARS-CoV-2 infection in the past. The briefing was supposed to share the findings of the survey.
The briefing may not even have happened if a breaking news story in The Telegraph (India) had not appeared on June 9. On the basis of information anonymously shared by public health experts, including some ICMR officials, the story said that the survey had found that in big cities the true level of infection may be 100-200 times the confirmed cases reported and up to 30 per cent of the population in these cities may have been infected up to early May. Through a tweet the same day, the ICMR denied this, saying that the report was speculative as the full analysis of surveillance data was yet to be completed, and called a press briefing two days later.
Interestingly, Bhargava began his briefing by stating that the numbers of confirmed cases and deaths per lakh population in India were among the lowest in the world. This was obfuscation at its best. Having conducted a sero-survey which, as he stated in the briefing, gave you information about the actual percentage of the general population that had been infected, to say that the number of confirmed cases (and not the level of true infection) was among the lowest in the world does not wash.
According to him, the sero-survey had two parts. The first was to estimate the fraction that had been infected in the general population and it had been completed. The second part was to estimate the fraction of the population that had been infected in the containment zones of hotspot cities, and that was yet to be completed. The survey also provided information on who were at a higher risk of infection and which containment areas needed to be strengthened, he said.
The districts selected for the sero-survey had been put into four categories based on the COVID cases reported on April 25—zero cases, low incidence, medium incidence and high incidence—and a minimum of 15 districts were selected in each category. In the table shown in the briefing (Fig. 2), even basic arithmetic is wrong: the total number of districts chosen is stated as 83, whereas it is only 71 (even this is two more than the 69 selected districts announced on May 12); the clusters selected add up to 710, whereas the table shows the number as 770; and the number of urban clusters shown as a percentage of the total number of clusters in each category seems to have been calculated wrongly. The data of 65 of these 71 districts had apparently been completed.
In the entire presentation neither the categories of low, medium and high incidence were defined nor the names of the 65 districts and the districts in each category were shared. This is the level of attention and care that the apex medical research body of the country gives to sharing information with the public on a pandemic, which reflects its attitude towards public information.
The following results of the survey were given in the presentation:
0.73 per cent of the population in these districts had evidence of past exposure to SARS-CoV-2;
Infection fatality rate (IFR) is low at 0.08 per cent;
The risk of infection in urban areas and urban slums was 1.09 and 1.89 times respectively that in rural areas; and,
Infection in containment zones (which containment zones?) was found to be high with significant variations, and this survey was still ongoing.
The conclusions that the ICMR drew from the survey were apparently the following:
The lockdown and containment had been successful in keeping the infection rate low and in preventing a rapid spread of the pandemic;
However, a large population still remained susceptible.
The first is clearly meant to show the policies being followed by the government to contain COVID-19 in good light when it is as clear as daylight that, as a statement issued by the Progressive Medicos and Scientists Forum (PMSF) about the briefing and data says, “the presented data seek more to hide the reality than reveal it.”
If you extrapolate the above data to the entire population of the country of about 1.4 billion, the number of infections (in early May) would be about 10.2 million, or more than one crore, whereas the actual number of confirmed cases in early May was 25,000-40,000 – a factor of about 300 down. In an earlier article (“Data Discrepancy”, May 8) Frontline had pointed out that the infection detection rate in April was about a factor 35 down. A factor 300 down now would imply that infections have increased greatly.
As regards mortality, even if one assumes that the IFR of 0.08 per cent, found as a result of the sero-survey, is correct for the total number of infections of 10.2 million, it works out to over 8,000 deaths, whereas data of early May showed 1,200-1,500 deaths. Does it mean that even deaths were being missed in data collection? Or, conversely, assuming that the number of COVID-19 deaths recorded was closer to the true figure as it would be difficult to hide deaths, if you use the IFR value of 0.08 per cent on the number of deaths on any given date, it should give the total number of infections. If you use it on the early May figure of about 1,500 deaths, nearly 1.9 million people had been infected in early May. Whichever way you look at the sero-survey data, they do not gel with the official data being released day after day.
That is why, very carefully, Bhargava chose the phrase “in these districts” (italicised above) while presenting the results of the survey. A sero-survey is supposed to give a projection of prevalence in the general population outside the districts/population sampled for the survey. Now, to give the benefit of the doubt to the ICMR, even if one restricted the projection to the 60 plus districts surveyed (about a tenth of the total number of districts), a rough calculation (by keeping out the urban districts from the original list of 69 and using the population data of 2011), the total number of infections worked out to over 0.35 million or 3.5 lakhs, an order of magnitude higher than the actual number of confirmed cases across all the 649 districts of the country. Even this geographically restricted projection reflects very poorly on the testing or infection detection rate.
A news story in the June 10 issue of The Economic Times , a day before the briefing, quoted D.C.S. Reddy, head of the ICMR’s 12-member panel of epidemiologists, as saying that the findings of the data had not been shared with the ICMR’s epidemiology and surveillance working group. However, The Economic Times had earlier reported that the ICMR chief had shared the survey data with the Cabinet Secretary on June 4 and with Health Ministry officials in a presentation via a videoconference on June 6. This would seem to be in line with the government’s present practice of keeping the scientific community out of policy decision-making or priority actions that need to be taken during a pandemic.
The PMSF statement further said: “The above claims are even more risible since it was admitted by the spokespersons themselves that the data from the COVID hot spots were still being compiled. The data presented represents the COVID situation existing in the country as of 30 April. In a rapidly evolving pandemic situation where the slope of the epidemic curve has steepened even further after the lockdown was lifted considerably, this data has already become dated in the 6 weeks gone by since then. It is noteworthy that in the countries, with which comparison has been done, the epidemic curve had flattened or was even falling after the lockdown…. These fallacious arguments seemed to have been forwarded to ultimately claim that there is no community transmission as of date in the country, hence, the need is to persist with containment strategy.”
A May 14 editorial in the journal Nature Cancer, titled “The tightrope of science, media and politics”, said, “The urgent need to address COVID-19 has highlighted the delicate relationships among science, politics and the media. To achieve a successful long-term response to the pandemic, stakeholders need to be guided by data, integrity and a sense of responsibility toward the public .”
It went on to say: “Some of these conflicts have become apparent during the response of different countries to the pandemic. Several European and Asia-Pacific countries made science-driven decisions to stem the spread of infection, established preventative measures, such as social distancing and use of face masks by the general public, and implemented widespread testing and contact tracing. Some of these countries are now reporting results that point to success in containing the virus… It is the responsibility of scientists to help shape public opinion on the basis of reliable knowledge. Scientists today are called to tread the delicate line of building public trust while providing sober assessments of data and managing expectations…maintaining an honest working balance with politics and media is also crucial, given that science does not operate in a vacuum…Now more than ever we need to remember that, similar to pandemic cures, change in public opinion does not come about with miracles, but through dialogue, respect and incremental but decisive steps .” (Emphasis added.)
It is the italicised part above that seems to be totally absent in the way the government and senior health officials have been taking policy decisions and implementing actions to contain the COVID-19 spread in the country. As of June 20, the number of confirmed cases and deaths increased by about 14,500 and 400 respectively a day. This, both from the above discussion and a very low rate of testing which even today is at just about 4,000 per million population, is far smaller than the actual figures. With the pandemic still very much at play and with no clear-headed strategy to handle the crisis evident from the policy-makers who have decidedly marginalised the scientific community, it is far from clear where the country is headed.
COMMents
COMMents
SHARE