Where WHO went wrong on India’s Covid death count

Under the law, the Chief Registrar of every state gets one year to send the compiled data on births and deaths to the RGI for collation and publication. CRS 2020 came out on May 3 and a two-year time lag has been standard even in the pre-Covid years. The cause of death is registered under the Medical Certification of the Cause of Death (MCCD) system — the scope of this data is limited because it applies only to urban areas and notified hospitals. The RGI is yet to publish it. But for the WHO and other mathematical modellers to surmise that India does not have a system to enumerate the dead and plunge into calculating mortality based on minuscule samples collected through informal channels is preposterous.

Death data from CRS shows an over 6 per cent increase overall in Maharashtra, Bihar, Gujarat, West Bengal, Andhra Pradesh, Tamil Nadu, Karnataka, Madhya Pradesh, Rajasthan, Assam, Andhra Pradesh, and Haryana. The CRS data even provides a breakup by age cohorts and comparing deaths among the elderly from 2018 to 2020 shows a reduction of deaths in the 70+ age group in Kerala and partially in Telangana. Of course, greater analysis and commentary are required but no one can say the system is dysfunctional.

An even more reliable tool than the CRS is the Sample Registration System (SRS) — it’s the largest demographic survey in the world and has been implemented since the Sixties. Eight million house visits result in a corpus of data on fertility and mortality. Such visits, however, were delayed during the pandemic, from 2020 to early 2022, when families, enumerators and reviewers were confronted with repeated Covid outbreaks and lockdowns. The SRS has been operational for decades although the priority accorded to timely data collection could certainly improve.

Under the law, the informer and the registrar must complete the process of registration within 21 days. Registration after 21 days requires approval at the higher levels — these seem to be progressively increasing — specified in the RBD Act. Surprisingly, Maharashtra and Delhi are missing from the relevant tables because “only partial information was provided”. This shows that the state responses are not accepted mechanically. The gaps and delays call for assigning higher priority and oversight to birth and death data collection.

The RGI has a statutory duty to declare births and deaths. For the calendar year 2020, the CRS figure is about 81 lakh deaths — this is 4.7 lakh more than the previous year. The SRS publication is expected soon which will, in all probability, show a higher figure. But in no case will it be anywhere near the 47 lakh deaths the WHO modellers have put out for 2020 and 2021 for India. The WHO’s Technical Advisory Group selected India for its modelling exercise without engaging with the RGI. Instead, modellers and academics have published a slew of papers using data like deaths among MLAs and Indian Railway personnel or based their studies on state websites or RTI responses to media queries. None of these is representative enough to justify extrapolation to cover a country as large and diverse as India.

India’s population is soon to be (if it is not already) the largest in the world. Tamil Nadu’s population is about that of France and Bihar is the size of Germany. The population of Uttar Pradesh, India’s most populous state, is about that of Brazil. In such a scenario, to base a WHO international report and academic research on sundry samples and extrapolate data to give alarming results is a disservice to a time-honoured national system. But it also overturns a convention that UN systems rely only on national data sources. The report has needlessly alarmed Indian citizens and the world community.

That tens of thousands lost their lives under tragic conditions that no one was prepared for, is a fact. Currently, around seven lakh applicants have applied for compensation related to Covid deaths — not millions. The publication of CRS reports has always taken two years which, without question, requires reduction. But that does not justify the WHO’s ferreting for fresh sources, almost implying that the state data is unreliable. If that were the case, women’s mortality data, for example, would not be showing a 40 per cent registration as against the 60 per cent registration for men — this is in consonance with the ground realities because women’s deaths are not viewed as material to matters like inheritance and property rights.

The good news is that awareness about the need for birth and death registration is growing. The RGI’s website indicates that the tightening of reporting schedules and making it mandatory for medical institutions and practitioners to report the cause of deaths is on the cards through amendments in the RBD Act. Digitisation at the primary level has touched 85 per cent and this will slash the data publication time. The use of digital software is increasing across states and dependence on manual reporting is reducing. Future RGI reports must be published promptly because only such data can alert the political system, policymakers and society of progress or the lack of it. This is what ought to stir reform.

India faces several challenges, but data collection and publication are not among them.

The writer is a former secretary in the Ministry of Health. Views are personal