The details of the 2 million excess deaths in 2021 | Number Theory

.

HT exclusively reported the data released by the home ministry that India recorded almost 2 million more deaths in 2021 – the deadliest year of the Covid-19 pandemic in India – than the preceding few years. Data from another report also showed that 81,000 more deaths were eventually attributed to Covid-19 than was first reported in the official pandemic bulletins issued by states.

People in PPE kits during the Covid-19 pandemic in 2021. (AFP)

There are two strands that stand out in the data: one, many states undercounted Covid deaths, perhaps because their health systems were overwhelmed; and two (and this is the more prominent strand), there was an inexplicable spike in overall deaths in 2021, by around 2 million.

It is difficult to understand how 2 million deaths could just not be counted -- unless the official data is wrong at some level -- and so, instead of attempting to answer that question, HT reviewed all three reports for 2021 released by the Registrar General of India – Civil Registration System (CRS), Sample Registration System (SRS) and Medically Certified Cause of Death (MCCD) – and answered three questions.

The details of the 2 million excess deaths in 2021

• 
  Were there states that undercounted even verifiable pandemic deaths?

  Yes. A comparison of state-wise Covid-19 death toll from the 2021 MCCD report and HT’s Covid-19 dashboard, which was built using official bulletins, shows that at least 14 large states undercounted pandemic deaths in their bulletins. The MCCD is not helpful in calculating the relative degree of undercounting across states because the proportion of registered deaths that are medically certified varies drastically across states. For example, only 4.5% registered deaths were medically certified in Bihar compared to 57% in Delhi. But clearly, the real discrepancy is in the spike in overall number of deaths – as seen in CRS and SRS data – in the country in 2021, which is difficult to explain unless one takes the pandemic into account.
• 
  Which states recorded the biggest undercounting of deaths in CRS and SRS?

  As HT explained on May 8, the CRS records all registered deaths and births in the country. The SRS, on the other hand, estimates deaths and births on the basis of a sample survey. Since not all deaths are registered, the SRS numbers on both births and deaths are generally higher than those recorded in the CRS. Gujarat, Madhya Pradesh, and Bihar are among the top five states with the highest level of undercounting of 2021 deaths in both SRS and CRS estimates among major states. On the other hand, Kerala, Maharashtra, and Delhi are at the bottom of the table in both. To be sure, because the level of registration of deaths in the CRS can show a large variation between states, the SRS is more suited for an inter-state comparison of the level of undercounting of deaths. SRS estimates show that Delhi, Gujarat, Haryana, Chhattisgarh, and Maharashtra had the highest jump in total deaths in 2021 if one were to compare them with the average deaths in 2017-2019 . Excess deaths as a proportion of 2017-19 deaths were the smallest in Telangana, Assam, Uttar Pradesh, Uttarakhand, and Rajasthan.
• 
  How much underreporting was due to Covid-19 patients not being able to seek health care?

  Anecdotal accounts at the peak of the second wave of the Covid-19 pandemic in May 2021 suggested this, as people queued up outside hospitals for beds and oxygen . At the time, experts also pointed to more deaths from other medical conditions -- simply for want of medical care. CRS data reinforces both . The share of deaths without medical attention increased from about one-third (34.5%) in 2019 to about half (47.3%) in 2021. To be sure, this number shot up even in 2020, when 45% deaths took place without any medical attention. In fact, the lack of medical attention even extended to births. Compared to an average 81.2% institutional births during 2019, only 73.7% births in 2020 were institutional, a number than decreased further to 72.9% in 2021. To be sure, the comparison of deaths without medical attention is not exactly kosher because the 2019 data did not have this data from Bihar and Jharkhand ( the data from 2020 and 2021 has). Because the CRS does not publish state-wise data on deaths without medical attention, it is not possible to ascertain the lower or upward bias in the 2019 numbers vis-à-vis 2021.