What new govt data reveals on the extent of undercount of Covid-19 deaths in India

The data from the Civil Registration System (CRS), published on Wednesday, is the first official indication of the extent of undercounting of deaths that occurred during the pandemic.

On May 9, the official total pandemic death count on the Covid-19 dashboard of the Union Ministry of Health and Family Welfare was 5.33 lakh. Assessments by researchers have put the actual number at 4-12 times higher.

A World Health Organization report released in 2022 put the number of “excess” deaths in India in 2020 and 2021 at about 47 lakh, almost 10 times the official Indian figure of 4.8 lakh for those two years.

India has its own system of counting births and deaths every year, but much of the data for the period of the pandemic were delayed. Besides the data from the CRS, the Census of India this week also released data from another system, the Sample Registration System (SRS), which estimates the total number of births and deaths in the country — and not just those that are registered and captured in the CRS — for 2021.

The government also released the 2021 report on the Medical Certificate on Cause of Death (MCCD), which provides information on the causes of death for deaths that can be medically certified.

Together, the data provide the most credible estimate of the number of Covid-19 deaths in the country, although parts of the picture are still missing.

However, recent trends do suggest that the difference between the actual deaths and the ones getting registered is narrowing.

‘EXCESS’ DEATHS

The big number in CRS 2021: more than 1.02 crore deaths were registered in the country that year, which was 21 lakh more than the 81.15 lakh deaths that were registered in the previous year (2020), the first year of the pandemic.

SRS data show that the number of people who died in the country every year from 2007 to 2019 ranged from 81 lakh to 86 lakh — an annual average of 83.5 lakh deaths. The majority of these deaths are registered and captured in CRS. The registration rate has increased considerably in recent years — rising from 70% in 2013 to 77% in 2016 and 92% in 2019.

This number may have risen even higher in 2020 and 2021, but those data are not available. But the trend is clear: the difference between actual deaths and registered deaths is narrowing.

The more than 1.02 crore deaths registered in 2021 can thus be considered a close approximation of the actual number of deaths that year. That would be 18.75 lakh more than the ‘normal’ of 83.5 lakh deaths annually. In the absence of any other reason for this significant increase in deaths in 2021, the entirety of this ‘excess’ would be attributable to Covid-19.

Even after accounting for some backdated adjustments, the number of deaths from Covid-19 during that year would not increase from the official figure (of 3.32 lakh) to, say, 3.5 lakh.

The CRS 2021 figure would, therefore, suggest that India’s official Covid-19 death figures were an undercount by at least five times.

But deaths occurred in all the three pandemic years from 2020 to 2022. The CRS data for 2022 are not yet available. What happened in 2020?

In 2020, the first year of the pandemic, the CRS registered 81.15 lakh deaths. Assuming that the death registration rate that year was the same as that in the previous year — 92% — the actual number of deaths in 2020 was about 88.21 lakh. This would be about 4.7 lakh more deaths than the ‘normal’ of 83.5 lakh.

Officially, India recorded 1.48 lakh Covid-19 deaths in 2020, or about three times less than the ‘excess’ deaths in that year.

There is more evidence for the extent of the undercount in the annual MCCD report for 2021.

Only about 23% of registered deaths in the country are medically certified. But even in the limited cohort of about 23.95 lakh medically certified deaths in 2021, more than 20% — 4.13 lakh — were attributed to Covid-19.

If all 1.02 crore deaths could be medically certified, this number would be bigger.

THE TAKEAWAY

The new data present very credible evidence for a more realistic assessment of deaths caused by Covid-19 in India. For these numbers are actual counts — and not estimates or extrapolations.

It must be nonetheless flagged that the calculations of excess deaths above are somewhat simplistic — and ignore factors such as the year-on-year variations in deaths, and the steady increase in registration rates.

They also do not account for the likely increase in deaths due to the disruptions that the pandemic triggered — such as the difficulties in accessing medical care — and not directly by the novel coronavirus.

Even so, the data released this week leave no doubt that India’s Covid-19 death count was in the range of at least 20 lakh, and not 5 lakh. The data also mean that the other studies and assessments, including the WHO’s, were probably a significant overestimate.

The ‘excess’ deaths in the country, for the years 2020 and 2021, calculated using CRS data in the relatively simple manner as above, were likely in the range of 21-22 lakh and not 47 lakh.

A crucial piece of data from the SRS, which would be available earlier, has been withheld from the 2020 and 2021 reports.

This is the estimated number of total births and deaths in the country in any given year. These numbers would be referenced in the CRS reports as well, along with the calculated registration rates for births and deaths.

These baseline numbers have not been mentioned in the reports — even though the SRS uses these numbers to calculate the percentages and other figures mentioned throughout the report.

The missing numbers can provide a more robust estimation of Covid-19 deaths in the country. For example, should it emerge later that the registration rate of deaths in 2021 was only about 75% or 80%, and not, say, 92% or 95%, the Covid-19 death count could get even bigger.

That said, however, registration rates have followed a consistently upward trend over the years, and it is unlikely to have dipped significantly in 2021.