Most people who die in Gadchiroli die at home. The remote, heavily forested district in central India is among the country’s least developed areas, and reaching the nearest clinic can take several hours on foot. After a death, relatives often bury or cremate their loved one in the fields.
When a COVID-19 surge arrived in India in 2020, it barely reached Gadchiroli. But in April 2021, as a devastating second wave tore through rural India, people began getting sick.
Yogesh Kalkonde, a public-health doctor and researcher who worked with the nongovernmental organization Search in Gadchiroli until July, says he soon began hearing about deaths in villages—sometimes four or five in a single small community. People seriously ill with COVID-19 began turning up at his rural hospital. But, Kalkonde says, he had little way of gauging the extent of the outbreak.
Similar situations were playing out in other parts of India. While official figures recorded averages of about 3,000 or 4,000 COVID-19 deaths per day, analysts saw signs that mortality across the country was far higher. But they struggled to produce even rough estimates of the true toll. Indeed, experts say that the pandemic has highlighted a long-standing issue: In the world’s second-most populous country, policy makers have historically paid too little attention to tracking people’s deaths, with serious implications for public health.
In fact, more than 100 countries do not have functioning civil-registration and vital-statistics systems that record births and deaths, according to the New York City–based public-health organization Vital Strategies. Several African countries collect little or no data on deaths. And the World Health Organization estimates that, globally, two-thirds of deaths are never accounted for.
These issues are acute in India, which is home to about one in six of the world’s people, and where around three-quarters of deaths take place in rural areas like Gadchiroli. In 2018, government surveys suggest, India registered 86 percent of all its deaths in its civil-registration system. Even then, many of the causes remain a mystery: Only one in five of those registered deaths was medically certified by a physician.
Those gaps, experts say, hampered public-health efforts in India and other countries long before the arrival of COVID-19. Governments can use death data for all sorts of public-health decisions: to identify malnutrition hot spots, address infant mortality, and even prioritize the shipment of vaccines.
“Counting the dead helps the living,” says Prabhat Jha, an epidemiologist at the University of Toronto and the director of the Center for Global Health Research, a not-for-profit organization co-sponsored by the university and St. Michael’s Hospital. “The main benefit of having data on who dies, and when, is to be able to understand what can be done about it today.”
Though counting the dead is the first step, finding out how they died is also key. Cause-of-death data are like the “thermometer of a health system,” says Kalkonde. Without such data, he adds, it is very challenging to track the progress of the health system.
Experts say the obstacles to improving data collection in India are steep. And even when such data exist, that doesn’t mean they will be accurate and available to the public—or to health researchers. The public-health doctor and researcher Sylvia Karpagam says she doesn’t expect much government effort to discover and publicize the real toll of COVID-19.
“Right now, it is all about making the government look good, and it is more about PR,” she says. The country’s leaders, she adds, “wouldn’t want to seriously look at how many people died.”
The government’s Ministry of Health and Family Welfare, refuting such criticism from activists and the media, said in an official statement in August that the suggestion that it was “missing out on deaths is completely unlikely.” The government’s Office of the Registrar General and Census Commissioner did not respond to requests for comment.
Each year in India, according to government estimates, 9.7 million people die—more than the entire population of New York City. Those deaths are scattered across a country with 23 officially recognized languages, nearly 600,000 villages, and a severe shortage of medical personnel. Keeping track of those losses in areas without a functioning health-care system is a difficult, but not insurmountable, task.
In large cities like Mumbai or Delhi, most deaths occur at or en route to a health facility. If a physician is present at the time of death, they fill out a cause-of-death certificate. Local authorities collect this information and pass it on to the state and then national government.
Far fewer deaths are certified in rural areas like Gadchiroli, where health facilities are bare or virtually nonexistent. Complicating the data collection, rural people often have little incentive to report deaths. Birth documentation is an important part of one’s legal identity, enabling access to services like education. But for those with few resources, living in places where informal transfer of property is the norm, there’s little reason to report deaths to authorities, especially of babies.
Gender and social status also affect which deaths are counted: Women, for example, are more likely to die at home, and less likely to have their deaths counted and certified, according to Kalkonde. And people who belong to India’s tribal communities, or who fall low in the Hindu caste hierarchy, typically have less access to health care and so show up less frequently in official statistics.
Then there is the dubious quality of the data. Not all doctors are well trained to certify deaths, says Kalkonde. “What you write on [the] death certificate in itself is a science,” he says. “So, even amongst the death certificates that are available, many times the cause of death is just listed as ‘cardiorespiratory arrest’”—meaning the person’s heartbeat and breathing stopped. “I mean, most of us die by cardiorespiratory arrest,” Kalkonde adds.
To account for some of these gaps, in the 1960s the Indian government began sending surveyors door-to-door in thousands of areas across the country in order to get a representative sample of the full population. When a selected household reports a recent death, the surveyor gathers the history of illnesses in the family, asks about the deceased person’s symptoms, and writes a report. Trained physicians look over the record and assign it a probable cause of death. Research suggests that this time-consuming process, called a verbal autopsy, provides a fairly accurate picture of the deaths. (China, the world’s most populous country, uses a similar system.)
In 2002, Jha and several collaborators began tracking mortality in 2.4 million households in India, using verbal autopsies to assign causes for deaths going back to 1998. The Million Death Study, as it is called, was far larger than other surveys, and its findings helped provide crucial insight into deaths by suicide, infant and child mortality, cancer in India, and more.
Still, researchers say, big gaps remain in the data. “It’s strange that, you know, 70 years down the road,” says Kalkonde, referring to the time since India’s independence, “we still have difficulty knowing what are the causes of death in India.”
Government officials aren’t always especially eager to publicize mortality data. Registering and certifying all deaths allows for accountability that leaders might not want, says Karpagam. And the findings of such studies don’t always align with state narratives about an accepting, egalitarian culture. Verbal autopsies have suggested, for instance, that one in every three women who die by suicide globally is an Indian woman.
In response, health officials and politicians sometimes hide or fudge statistics. Those problems have worsened since 2014, Jha says. Since then, there have been more delays in releasing India’s sample survey. Recent years have also seen the delays in farmer-death statistics.
When a massive second wave of the pandemic hit India in the spring of 2021, the shortcomings of these gaps in mortality data were suddenly on display, as there was hardly any accounting of how many people in India’s rural areas were dying from COVID-19. The data disaster made global headlines.
In response, some Indian states have begun conducting door-to-door surveys, trying to pin down more accurate figures. Others released revised death statistics after courts stepped in to flag irregularities.
The solution to the improving reporting of death data is different in urban and rural areas, Jha says. “In urban areas many deaths occur in hospitals, so you can improve that registration and reporting system,” he says. In rural areas, he adds, some of the villages already track deaths. “I think that’s important, but that needs to be catalogued and shared.”
A treacherous 10-hour drive along winding mountain roads northwest of Gadchiroli is the tribal region of Melghat. Deep in the forests, the phone network sometimes drops to zero. Melghat is known for its tiger reserve. It’s also poor: The region occasionally makes news for the high rate of child malnutrition deaths among its largely tribal population, spread across about 370 villages.
Mittali Sethi came here on her first government posting, part of a national cadre of civil servants called the Indian Administrative Service. Until transferring to a new position in July, Sethi, a trained dental surgeon, was in charge of the development of the tribal population, including overseeing health programs.
According to Sethi, there is one private physician in Melghat and a handful of government ones. For specialized medical care, she says, locals have to leave the district and travel about three hours in a vehicle to a facility in the nearest city, Amravati.
Here Sethi put into practice what Jha recommends in theory. Through a national program launched in 2005, India’s villages have an army of women community health-care workers called Accredited Social Health Activists, or simply ASHAs, which in Hindi translates to “hope.” Often working in their home villages, these women are trained in basic health-care delivery, like providing pregnant women with iron and folic-acid tablets, and helping with awareness campaigns on sanitation.
As the second COVID-19 wave arrived in April, Sethi reached out to the ASHAs in Melghat and asked the women to report every death to their nearest health center on a daily basis, along with the likely cause. “We were monitoring which village was having more deaths, because people were a little reluctant to test,” Sethi says. “So the proxy for knowing where COVID might actually be increasing was to see the number of deaths.”
These basic data were enough to give Sethi and her team information about where to step up COVID-19 awareness and vaccination campaigns. “Not having data has never been a problem here,” she says. The resultant campaigns were successful in countering vaccine hesitancy and made national news in India.
What Sethi did in Melghat is what Jha says could be done in every district in India. The approach essentially decentralizes the health system, gathering and responding to data in real time instead of waiting for a national-level analysis.
As other countries grapple with similar problems, experts are hoping the pandemic will draw attention to death registration. Having better data, the World Health Organization writes in a new report on global birth and death registration, would help communities around the world combat COVID-19—and forms “a cornerstone of a strong health system for the future.”
Romain Santon, the deputy director of civil registration and vital statistics for Asia at Vital Strategies, which has helped 29 countries bolster their civil-registration systems, says simple policy changes can yield better data. In some countries, he says, burial permits or other legal requirements could act as incentives to push up death registrations. Making registration simpler by reducing the number of forms, or digitizing the process, could also help.
Santon calls for compulsory, confidential, and permanent death registration worldwide. “Every single individual living in a territory,” he says, “should have access to birth and death registration.”