Data: Understanding the ‘Death Registration’ numbers of various states
Sai Krishna Muthyanolla
May 13, 2022
The recently released excess COVID-19 mortality estimates by the World Health Organization (WHO) attracted severe criticism from multiple countries, particularly India. To counter the excess mortality estimates by the WHO, the government relied on the recently published Vital Statistics of India based on the Civil Registration System 2020.
The recently released excess COVID-19 mortality estimates by the World Health Organization (WHO) attracted severe criticism from multiple countries, particularly India. The WHO estimated the COVID-19 linked deaths in India to be around 4.7 million, which is by far the highest for any country and makes up for nearly one-third of the global deaths. The official government death toll as of 12 May 2022 stands at 5,24,181, equivalent to approximately 0.5 million. The excess mortality comes around 0 times the official government data.
The Government of India (GoI) refuted the claims made by the WHO on excess covid mortality estimates. India’s primary objections were linked to the validity and reliability of the data models and the data collection methodology.  GoI had also highlighted the inverse relationship between temperature and mortality, as assumed in the data modelling, despite repeated requests to substantiate the same. GoI also questioned the usage of data modelling, when authentic data is available in the form of the Civil Registration System (CRS), through a robust network of entities under the Register General of India (RGI).
To counter the excess mortality estimates by the WHO, the government relied on the recently published Vital Statistics of India based on the Civil Registration System 2020. The government clarified that since this report publishes ‘all-cause mortality’, the mortality associated with COVID-19 can best be inferred to be a sub-set of this ‘all-cause mortality’ for that particular year. There are similar objections made by the government regarding the usage of test positivity rate, demographic, and size considerations in arriving at the estimation of excess COVID-19 deaths.
In this article, we look at a few issues surrounding the data on COVID-19 mortality to bring some clarity to the whole issue.
Indian death statistics
As stated earlier, CRS data represents the ‘all-cause mortality’ in India. The data is sourced from the annual statistical reports prepared by the respective chief registrars of the states and Union Territories. The period for the 2020 report is from 01 January 2020 to 31 December 2020.  The number of registered deaths in India in 2020 stood at 81,15,882, which was 76,41,076 in the year 2019. This represents an increase of 6.2% of the registered deaths in 2020, as compared to 2019.
The majority of the Indian states have shown an increase in the number of registered deaths in 2020 compared to 2019. States like Jharkhand, Kerala, Manipur, Telangana, Uttar Pradesh, Uttarakhand, and Delhi reported a decline in the number of registered deaths in 2020 as compared to 2019. To see if this is an anomaly, the trend from 2011 is plotted for these states. This is not a one-off incident, as these states reported a decline in deaths in past years as well. However, what makes this year’s decline significant is that the world has faced a deadly pandemic. At a time when countries across the globe, including India, reported a rise in deaths, the decline in the share of deaths for these states should be looked with careful consideration. It must also be kept in mind that the figures from the CRS data indicate only the registered cases and not the actual cases. Thus, the level of registration is critical to understanding the gap between registered and actual cases. Incidentally, the percentage of death registrations completed within the prescribed limit of 21 days is low for these states, with Jharkhand at 50.7%, Kerala at 55.3%, Telangana at 65%, Uttar Pradesh at 50%, and Uttarakhand at 44.2%. In fact, nearly 15 States and Union Territories had less than 80% of death registrations done within 21 days.
COVID-19 deaths as a percentage of registered deaths
The government, in its rebuttal, made clear that the COVID-19 deaths at best can be inferred as a subset of the total CRS reported mortality in India for that period. Accordingly, the COVID-19 deaths till the end of 2020 are considered to map with the CRS all-cause mortality in 2020. No state except Uttarakhand and Maharashtra accounted for more than 2.5% of the COVID-19 related deaths of the total registered deaths as per CRS in 2020. States like Uttar Pradesh, Telangana, Kerala, which reported a decline in all-cause mortality in 2020 compared to 2019, had less than 1% of their total registered deaths associated with COVID-19 in 2020.
Exploring the plausible reasons for the decline in deaths in certain states
As observed above, states like UP, Telangana, Kerala, Delhi reported a decline in the registered deaths in 2020. To understand the possible reasons behind such decline, we looked at the accidents and suicides cases in India, data of which is published by the National Crime Records Bureau (NCRB). This analysis is based on two specific assumptions:
Accordingly, the data is mapped for these states. It is found that except for Uttar Pradesh, almost all states followed a normal trajectory in suicides and accident-related deaths in 2020. Even in the case of Uttar Pradesh, this decline is not significant, when compared to its own overall decline in deaths. Hence, these figures also do not justify the decline in death cases for these states.
Life insurance claims as a proxy for COVID-19 deaths
Insurance claims can be considered as a representative metric for looking at the deaths that occurred. A Right to Information (RTI) by the Indian express reveals that the total COVID-19 death insurance claims settled by both the private insurers as well as the Life Insurance Corporation, from the beginning of the pandemic stand at 2.2 lakhs, amounting to Rs. 17, 362.48 crores. However, for the fiscal year 2020-21, the claims settled were only Rs. 1418.71 Crores, with around 21,304 claims.
The significant difference between the total settled amount from the beginning of the pandemic and the settled amount in 2020-21 indicates that a vast number of claim settlements occurred after 2020. This is because the second wave had a more devastating impact than the first wave. Another reason for the spike in the settlements could possibly be the improvement in the percentage of insurance penetration in India, which improved from 2.8% in 2019-20 to 3.2% in 2020-21. The value of claims rejected also rose sharply from Rs. 575 Crores in 2019-20 to Rs. 865 Crores in 2020-21. This could be due to the confusion in the initial days of the pandemic as to whether it is covered under insurance policies. Hence, unless we have detailed registered death figures for 2021 (CRS 2021), it is difficult to understand the reasons for such trends in some states.
Level of registration hidden in CRS 2020 report
As mentioned earlier, the level of registration is critical to understanding the gaps between actual deaths and registered deaths. CRS data has only registered deaths and not actual deaths. Every year, the CRS report gives the level of birth and death registration. Coincidentally, this year’s CRS did not make any mention of the level of registration. Instead, a press release through PIB mentioned the death registration rate for 2020 as 99.9% following the WHO’s excess mortality claims.
Observations on the ongoing debates on excess mortality
In conclusion, it is important to comprehend the fact that the idea behind any model of death estimates is to give a fair warning about future pandemics. Had the data regarding the death reporting system been transparent and accurate (remember the no-death due to lack of oxygen claims), these kinds of models would not have made headlines. It is high time that the death reporting system must be made robust, transparent, and up to date. In this age of dashboards & real-time data, it is not difficult for the government to make these statistics available more real-time than release a report months later.
Featured Image: WHO excess COVID-19 mortality estimates