Under-reporting of pandemic toll

The concealing of the actual numbers of the dead is likely to impact research on the COVID-19 pandemic and block efforts to combat the virus.

A mass cremation, at a crematorium in New Delhi in April, at the height of the second wave.

S cenes of mass pyres, burial grounds and bodies floating down the Ganga will be the most haunting images from the COVID-19 pandemic. On June 23, the official COVID death toll in India was 3.89 lakh, which was enough to place the country among the worst affected in the world. According to unofficial estimates, the real number would be in multiples of the official figure. Experts who track mortality data say deliberate under-reporting in several key States will only block efforts to combat the deadly virus.

As the country prepares for the third wave, Frontline spoke to people monitoring the numbers. If the surge in the death count is analysed, it will help the government understand basics such as the demographics of the infected, how the disease is spreading, and where the hotspots are and why. These are fundamental data, they say, that can go a long way in getting India out of the mess it is in.

Usha Ram is a doctorate researcher at the Department of Public Health and Mortality Studies, International Institute of Population Studies, Mumbai. She said: “Unfortunately, India’s death registration system is not as efficient as it should be. Unless we can get specific and real-time numbers, which is critical for handling COVID, we will not be able to move forward. The only source for COVID numbers in the public domain is the same source the government uses. It is at the aggregate level and not unit level. To understand a pandemic or the mortality scenario in the country, you should have your death data stratified by age and sex. Just the total number of deaths will give me absolutely nothing.”

She added: “The figures are definitely grossly under-reported and objectively no one can debate it. We cannot do anything about it as we have no control over data sources and its dissemination in India. Even RTI [Right to Information Act] does not give you the provision to ask for disclosure of cause of death. Till date we do not have a data source in India which can give reliable causes of death data, even at the State level. We have a good HMI [Health Management Information] system that has evolved over the past 15 years, but there also, unfortunately, the information recorded, reported and disseminated with regard to cause of death is not reliable.”

Excess mortality

Excess mortality is defined as an estimate of the additional number of deaths within a given time period in a geographical region, such as a country or a State, compared with the expected number of deaths, usually estimated using the same time period in the preceding year or the average over two or more years. A paper titled “Excess mortality: the gold standard in measuring the impact of COVID 19 worldwide?” published by the Journal of Royal Society of Medicine , United Kingdom, said: “Under the assumption that the incidence of other diseases remains steady over time, excess deaths can be viewed as those caused both directly and indirectly by COVID-19 and gives a summary measure of the ‘whole system’ impact. As the pandemic has progressed, there has been a growing focus on excess mortality as a more reliable metric for comparing countries.”

The issue of excess mortality in India came to the fore in early May when Divya Bhaskar , a Gujarati daily, collected death data through field investigations. Intrepid journalists ferreted out facts from crematoriums and burial sites, tracked the increasing number of obituaries in dailies, looked at issuance of death certificates, and positioned themselves outside hospital mortuaries where they could at first hand witness bodies being brought in. The mismatch between official figures and ground realities exposed Gujarat’s concealment of death numbers. It was one of the first States to be thus exposed. Subsequently, reports revealed Madhya Pradesh, Andhra Pradesh, Uttar Pradesh, Bihar and, more recently, Tamil Nadu to have under-reported numbers.

Divya Bhaskar ’s reports said that during a 71-day period (March 1 to May 10, 2021) as many as 2,891 deaths could be accounted for as having been caused by COVID in Bhavnagar. The Gujarat government’s official COVID dashboard said there were 279 deaths in Bhavnagar since the start of the pandemic. In Ahmedabad, which has been a hotspot for long, reporters found that from March 1 to May 10 in 2020, the city had issued 7,786 death certificates. In the same 71-day period in 2021, it issued 13,593 death certificates. The State government’s COVID portal reports 2,126 COVID-related deaths in this period. Divya Bhaskar compared these numbers to the death statistics of 2019 and 2018 and found an exponential jump in the death rate.

An epidemiologist in Mumbai said: “Obviously, the implications of excess mortality are significant. It blatantly reveals that the administration is failing. Yet, procuring the data is critical. This is really the only way to have a good sense of the toll in society.”

In India, the nationwide Civil Registration System (CRS) is considered the foremost source of records for births and deaths. The registration process, however, is not dynamic and the most recent data would be that of 2019. Led by the Office of the Registrar General of India, the data collection happens at the State level and is expected to record all deaths from all causes and all locations, whether medically certified or not. However, doctors say that in most States the cause of death is not recorded properly.

Data experts, demographers and media houses have been collaborating to analyse and publicise the staggering death toll witnessed in the country recently. Rukmini S., a data journalist based in Chennai, has relentlessly pursued the excess mortality issue. In her study on Madhya Pradesh, where she crunched numbers primarily from the CRS, Rukmini says the State reported 1.6 lakh deaths in May 2021 alone. In May 2018 and 2019, the average was 31,000, which means the excess was nearly five times the base line fatality. The official COVID toll was 4,461 in May 2021.

Although CRS has not published recent data, sources say researchers accessed internal CRS data via contacts in the system.

An independent article by the news website Scroll.in reports that the official number of COVID deaths in Andhra Pradesh in May 2021 was as low as 967. According to the writer’s sources, however, the total deaths in the State numbered a staggering 1.3 lakh that month. Even if all of these deaths were not caused by COVID, the excess mortality was nearly five times the average number of deaths (27,000) seen in May 2018-2019. Tamil Nadu only provides an annual total of registered deaths. The average for the combined years of 2018 and 2019 was 5,93,000 deaths. According to the writer’s sources, during a six-month period from January 2021 to June 2021, which included the months of the second wave of the pandemic, the State reported 4,00,000 deaths. Article 14, an advocacy website, says that following an RTI inquiry asking for the total number of deaths in Uttar Pradesh in the pre-pandemic months, 24 districts were found to have registered around 1,78,000 deaths from July 2019 to March 2020. The same districts recorded 3,75,000 deaths from July 2020 to March 2021, an increase of 110 per cent.

Understanding death

Aashish Gupta, a demographer and sociologist who has written extensively on the pandemic and excess mortality issues, said: “Measuring excess mortality reveals a truer picture and can help in preparing for the third or the next few waves that are inevitable. Typically, excess mortality occurs during a pandemic or an event such as a famine. It is otherwise rare to see such high mortality.” Gupta said that there were now three parallel systems for recording deaths: the Indian Council of Medical Research (ICMR) recording COVID deaths, the CRS, and there were State, district and panchayat-level registrations. “In spite of this, the numbers researchers and journalists are pulling out are astonishing,” he said, adding that the real picture would not emerge unless there was a cohesive system.

Dr Murad Banaji, a mathematician and public health expert from Middlesex University, U.K., who has written on India’s COVID situation, said: “As the second COVID surge was developing nationally, there was little or no mitigation in place. It is likely that high mobility and high levels of active infection coincided. This would have allowed the virus to spread widely and rapidly, including in rural areas, and cause high death tolls. This scenario likely played out in many States, including Gujarat—to find out more we’d need more local mortality data.”

He added: “More accurate data would be of great help in understanding the epidemic and predicting what might come next. But it is clear that some State and city administrations fear the political consequences of accurately reporting COVID-19 mortality. States such as Maharashtra that have reported high numbers of COVID-19 fatalities have effectively been described as failing by the Central government. This narrative is highly dangerous: it discourages transparency and encourages dishonesty.”

“I’m not sure if India has failed to invest in an efficient recording mechanism. But it is true that there are various weaknesses which contribute to under-reporting of COVID deaths. Even in normal times, many deaths are not medically certified, and some are not registered in the civil registration system. At the moment, deaths which do not occur in a hospital, or occur without a positive COVID-19 test are not, in general, recorded officially as COVID deaths. On top of this, the choice by some death audit committees to omit people with co-morbidities from official figures increases under-reporting,” he added.

Banaji said that the Gujarat Chief Minister’s misinterpretation of ICMR recommendations suggested that some of the under-reporting in Gujarat resulted from deliberate policy at the State government level. (Chief Minister Vijay Rupani had implied in an interview aired on Sandesh News that according to ICMR guidelines the deaths of COVID-19 patients with pre-existing conditions like cancer and diabetes should not be classed as COVID-19 deaths. The guidelines, however, are clear that deaths of COVID-19 patients with co-morbidities should, in general, be recorded in the official toll. Moreover, suspected COVID deaths should also be recorded, even when the patient does not have a positive COVID-19 test.) There are reports from Gujarat of patients who died in hospital after testing positive for COVID but were still omitted from official death counts.

Usha Ram said: “The suppression is obvious. The backlash will be huge. The issue of recording deaths is very complex.” If it can be done efficiently with unit details, it will have far-reaching results for a country’s health system. In advanced countries, as soon as a child is born, the information is recorded in a central system, which tracks every vaccination and health issue of the individual for the rest of her/his life. The data are made available to researchers, which when analysed can lead to major advancements in health care, she said.

The pandemic has highlighted a problem that was perhaps never considered important and has become deeply entrenched. Clearly, the dead can help. If only they were counted.