Staring into a black hole. Data on COVID-19 mortality in India.

Aug 7, 2021 | All posts, Country reports

Peter Lloyd-Sherlock, Aravinda Guntupali, Arokiasamy Perianayagam and Martin McKee.

7 August 2021

The Covid-19 pandemic has brought into sharp focus the lack of a comprehensive system of statistics on mortality and cause of death in India. By 25 May 2021, only 299,296 Covid-19 of an estimated 654,395 deaths were recorded officially [1]. Volunteers in the state of Kerala found records of 4,559 deaths in local media between March and December 2020, but only 2,646 deaths reported by official sources [2]. Similarly, 8,286 cremations were reported for seven Gujarati cities between 15th April and 3rd May 2021, compared to only 1,061 officially recorded deaths [3]. Just 50 Covid-19 deaths were reported for the city of Bhopal during the first three weeks of April 2021, but other sources estimated there had been over a thousand [4]. If anything, these ad hoc studies may understate the true level of under-reporting, since not all deaths are recorded in the media.

These large disparities should come as no surprise, given the failure to establish an effective civil registration system. As recently as 2015, less than a third of deaths in some Indian states were reported [5]. Of those recorded, only around 22 per cent receive a medically certified cause of death [6]. India’s limited mortality data contrasts with the good availability of data in many other countries [7]. Importantly, a study of completeness of death registration in 2015 found very low, and declining, rates in some Indian states, such as Nagaland (29%) and Manipur (34%), while others, from Kerala and Tamil Nadu in the South to Punjab and Haryana in the North, had achieved 100% coverage [5].

Indian data on all-cause mortality are too incomplete to permit comparisons over time. These are needed to capture the overall mortality impact of the pandemic, including likely increases in deaths from other conditions due to reduced access to treatment and other effects. There is also a problem of transparency. In 2020, Indian and global researchers asked the Government of India to release the civil registration data for deaths during the pandemic, along with mortality data for the two preceding years, in order to calculate excess deaths [8]. This request was not granted, perhaps reflecting the government’s own lack of confidence in these data.

As India’s data on all-cause mortality are so poor, more disaggregated analysis of infections, comorbidities, hospitalizations and fatalities by age, sex and socio-economic status are not possible. This is of great concern, since these data are necessary to inform policies such as vaccination priorities.

We make three urgent recommendations.

First, we urge organisations that compile comparative international datasets on COVID-19 mortality, including John Hopkins and WHO, to be more careful in their presentation of data for countries like India, where their quality are so poor that they do more to obscure than reflect the true impact of the pandemic. This could include clearly signalling cases where the available data appear to be highly unreliable.

Second, we urge India’s national government to commit the required investment and support for a more complete system of death registration. This should be done in conjunction with the strengthening of primary health care facilities in rural areas, which can serve as registration and surveillance points for deaths that occur at home, and should ensure learning from those states that have achieved high registration rates. To enhance cause of death attribution, the use of verbal autopsy by primary health care staff should be scaled up markedly [9].

Third, as a short-term expedient, we recommend the rapid extension and adaptation of existing demographic and health surveys to capture mortality at all ages, as well as other direct and indirect effects of the pandemic. This will provide useful insights about high-risk populations in deprived settings.

 


REFERENCES

  1. COVID Kerala 2021 covid kerala – Google Sheets
  2. Institute for Health Metrics and Evaluation (IHME) 2021. Estimation of total mortality due to COVID-19 http://www.healthdata.org/print/8660
  3. The Quint (2011). Local Newspapers Question Gujarat Govt’s Data on COVID-19 Deaths (thequint.com)
  4. The Wire Science 2021. The Importance of Knowing How Many Have Died of COVID-19 in India – The Wire Science
  5. Kumar GA, Dandona L, Dandona R. Completeness of death registration in the Civil Registration System, India (2005 to 2015). Indian J Med Res. 2019;149(6):740-747.
  6. Chatterjee P. Is India missing COVID-19 deaths? Lancet 396(357) September 5, 2020.
  7. United Nations Statistics Division. Civil registration and vital statistics. https://unstats.un.org/unsd/demographic-social/crvs/#coverage
  8. Appeal for excess death data, 2020. pdf – Google Drive
  9. World Health Organisation (WHO) (2016) Verbal autopsy standards: ascertaining and attributing causes of death. https://www.who.int/healthinfo/statistics/verbalautopsystandards/en/