Home / Opinion / Columns /  What they don’t tell you about India’s official death toll

The latest estimates of the pandemic’s toll in India from the World Health Organization (WHO) suggest that the health crisis claimed nearly 5 million lives in the country over the past two years. The WHO estimate of the pandemic toll includes both the direct and indirect impact of covid. It includes deaths of non-covid patients who died because of lack of access to timely care, for instance, when hospitals were filled with covid patients and had to turn away others.

Official spokespersons of the government have raised objections against WHO estimates, and even questioned the motivations of its team. Citing the recently-released civil registration system (CRS) report for 2020, the Centre has claimed that the count of extra deaths in 2020 was actually much lower.

These objections don’t make sense. The CRS report suggests 0.8 million excess deaths in 2020 compared to the average of the two previous years. The WHO data also suggests a similar number of excess deaths for 2020. WHO’s final tally for the 2020-2021 period is much higher because the WHO estimates nearly 4 million excess deaths in 2021. The CRS report for 2021 has not been published yet and hence a direct comparison is impossible. But provisional CRS figures for several states accessed by journalists (and which the WHO used in its estimation) suggest that the summer of 2021 (when our second wave of covid peaked) saw much higher death registrations than in 2020.

The final Indian official tally based on CRS is likely to be close to the WHO estimates, since it would include the second wave’s impact. A convergence between the Centre’s death tally and WHO figure, once 2021 numbers are taken into account, should not surprise anyone, since they both rely on the CRS.

This raises the question of whether the CRS tally is an accurate measure of deaths in India. The answer, unfortunately, is a big ‘no’. The CRS heavily undercounts deaths in some of India’s most populous states, such as Bihar and Uttar Pradesh. It also undercounts deaths of children and women.

The Registrar General of India (RGI), which runs the CRS, itself doesn’t regard the CRS as the most authentic barometer of mortality across the country. It uses a randomized large-scale survey, the Sample Registration System (SRS), to estimate death counts. The ratio of registered deaths (as recorded by the CRS) to estimated deaths (from the SRS) is defined as the “level of registration", which indicates the coverage of the CRS system.

This system worked fine as long as the SRS was able to measure deaths accurately. But the dark secret of India’s official death-recording system is that SRS data quality has slipped over time. Independent studies show that the SRS has been under-counting deaths especially in better developed states. A 2010 study by Prasanta Mahapatra of the Hyderabad-based Institute of Health Systems suggested that the SRS under-count is higher in states like Andhra Pradesh and Maharashtra than in states such as Odisha and Uttar Pradesh. A 2020 study by Chalapati Rao of the Australian National University showed that the SRS undercounts adult male deaths in most states.

The expansion in the CRS over the past decade exposed fatal flaws in SRS data. In states such as Haryana, Karnataka and Tamil Nadu, the CRS now reports more deaths than the SRS, meaning that the level of registration technically exceeds 100%. Yet, for children and women, the CRS still reports fewer deaths compared to the SRS in most states.

So we have a uniquely broken death-recording apparatus in the country. The CRS fails to capture mortality among the poor, among women and among children. The SRS fails to capture deaths among the non-poor and adult men.

Given that both the systems are deeply flawed, it is impossible to deduce the true level of registration from the CRS and SRS numbers. A more reliable measure comes from the National Family Health Survey (NFHS), a large-scale household survey that is nationally representative. The latest numbers for 2019-21 show that just 71% of households had registered deaths in their family in the three years preceding the survey. For the poorest wealth quintile, this figure was even lower at 52%. Among major states, Bihar (36%), Jharkhand (40%), and Uttar Pradesh (47%) saw the lowest levels of registration. Goa was the only state which reported complete death registration.

Rather than fight pointless battles with WHO technocrats, Indian policymakers need to find ways to revamp our death-recording apparatus. CRS coverage has expanded over time, but it is well short of universal coverage. Estimates of the pandemic’s toll based on either CRS or SRS data would therefore miss numerous deaths in many parts of the country. Such estimates can only provide a lower bound for the true toll of the pandemic.

It is time to redouble efforts to improve the CRS and SRS systems, since they are both extremely important to understand mortality and health patterns across the country. The database on ‘cause of death’ also remains poor. This too needs urgent attention.

Rather than harangue global experts for doing their job, Indian policymakers should engage with them to help improve our vital statistics. Unless we are able to obtain accurate patterns of deaths and diseases in the country, we will always struggle to improve health outcomes. And unless such data is available on a real-time basis, we will remain under-prepared for the next health shock. We should not need any prodding from multilateral agencies to publish routine health statistics in good time.

Pramit Bhattacharya is a Chennai-based journalist. His twitter handle is pramit_b

Subscribe to Mint Newsletters
* Enter a valid email
* Thank you for subscribing to our newsletter.
Recommended For You
Edit Profile
Get alerts on WhatsApp
Set Preferences My ReadsFeedbackRedeem a Gift CardLogout