In April 2021, Kailash Tiwari lost his previously healthy 100-year-old father in India to COVID-19.
Soon after that, two close family members aged in their 60s died with the virus.
Several of his friends, colleagues and professors also lost their lives during the country's severe Delta variant outbreak.
"Each and every house had at least one or two people that died," Mr Tiwari told the ABC.
Mr Tiwari left his home in Adelaide to visit India shortly before the country was struck by a catastrophic COVID-19 wave between April and June of last year.
He was able to enjoy walks with his father in a small village of 50 people in Indore before nearly half of the residents were infected with the virus.
India's Delta wave sparked a humanitarian crisis created by an acute shortage of hospital beds, medicines and oxygen.
Amid a deluge of deaths, Delhi residents reported having to keep the bodies of loved ones at home for days.
Mr Tiwari's family was able to give his father a proper burial, but for many others, that was not the case.
At the height of the outbreak, workers at makeshift facilities were carrying out mass burials and cremations.
Aerial photos showed funeral pyres just metres apart in packed open crematoriums and bodies burning atop piles of wood in car parks.
Despite the clear signs of the large-scale impact of the virus, the country's official COVID-19 death toll sits at about 520,000.
But this week the Indian government said there had been 475,000 excess deaths (out of 8.1 million total deaths) in 2020 compared to the previous year.
Vinod Kumar Paul, a top health official who has overseen India's fight against the pandemic, said there was nothing "dramatic" in the total death data for 2020 and the numbers were "absolute, correct and counted".
India's figures were released ahead of the World Health Organization (WHO) report on global excess death estimates, which New Delhi openly opposes due to what it says are its "questionable" methods.
Thursday's report — the most comprehensive look at the true global toll of the pandemic so far — said that globally there were 14.9 million excess deaths associated with COVID-19 by the end of 2021.
The new figure from the WHO is more than double the official reported tally of 5.4 million deaths.
The report suggests that 4.7 million people died in India between January 2020 and December 2021 as a result of the pandemic, mainly during a huge surge in May and June last year.
It therefore attributes almost half of the deaths that until now had not been counted to India.
The WHO's excess mortality figures reflect people who died of COVID as well as those who died as an indirect result of the outbreak, including people who could not access health care for other conditions when systems were overwhelmed during huge waves of infection.
So how can data from the Indian government and the WHO differ so dramatically?
Under-counting not 'active suppression or manipulation'
Since the start of the pandemic, experts have believed that official government figures have been significantly under-reported.
Rukmini S, an independent data journalist from Chennai, India, has been closely tracking and documenting the country's COVID-19 impact.
It was not a surprise to her that not all deaths were being counted.
"India systematically under-counts its deaths from other diseases as well," she told the ABC.
"In rural areas or among more marginalised communities, some deaths just never make it to [the] official record. So they don't get registered."
When COVID-19 struck, many states adopted different definitions of what it was, and a lack of testing meant some people died without being diagnosed.
"But none of this was active suppression or manipulation," Rukmini said.
As the second wave hit and cities were "literally running out of space to cremate bodies", it was clear that the officially reported figures were not matching those numbers, so local journalists and researchers began gathering their own data.
They accessed state-level Civil Registration System (CRS) portals that count all deaths in a district for every month and were able to see an "enormous spike" that wasn't being reflected in the official toll.
"All of this seems to be in a sense validated by the WHO estimates," Rukmini said.
Other independent assessments have put the death toll in India far higher than the government tally, including a report published in the journal Science in January which suggested 3 million people may have died of COVID-19 in the country.
Another report from July last year — published by Arvind Subramanian, the Indian government's former chief economic adviser, and researchers at the not-for-profit Center for Global Development — also put deaths between 3.4 million to 4.7 million.
It said the official count could have missed deaths that occurred in overwhelmed hospitals or while health care was disrupted.
"True deaths are likely to be in the several millions, not hundreds of thousands, making this arguably India's worst human tragedy since Partition and independence," the report said.
What exactly is India opposing?
The WHO process, which involved a panel of global experts convened jointly by the WHO and the United Nations Department of Economic and Social Affairs (UN DESA), began in February 2021.
The panel used a combination of national and local information, as well as statistical models, to estimate totals where the data was incomplete — a methodology India criticised.
Since at least December 2021, the Indian government had been stalling the WHO's process by refusing to share data, Rukmini said.
"The government systematically pushed back … it promised to produce estimates of its own and did not. It has only now produced its annual estimates for 2020 on Tuesday — 48 hours before the WHO's scheduled release of its data."
Rukmini said she was informed by the WHO that India was the only country that had these objections.
The WHO said it had not yet fully examined new data provided this week by India and might add a disclaimer to the report highlighting the ongoing conversation with the country.
Much of the data used by the WHO comes from 18 states which represent 70 per cent of the Indian population.
Despite the data coming from state-level portals, India argues it was "obtained from some websites and media reports and was used in their mathematical model".
"This reflects a statistically unsound and scientifically questionable methodology of data collection for making excess mortality projections in case of India," the health ministry said in a statement after the WHO's report was released.
The government said only data from the Registrar General of India based on the Civil Registration System should be used for such estimations.
It has repeatedly defended its death registration system as being "robust", saying: "Missing out on deaths is completely unlikely."
India has also rejected its classification as a "Tier II" country in terms of data availability.
The WHO report classified countries into three tiers, which it said were statistical, not developmental, classifications.
Steve MacFeely, director of data and analytics at the WHO, told an online forum last month that some countries had been upset by their tier placement.
"This doesn't say anything about the quality of the data in their country; what it talks about is the availability of that data," he said.
Why is the new data important?
More than 60 per cent of countries don't register their deaths, Mr MacFeely said.
Only 39 countries around the world produced what the WHO considered to be high-quality cause-of-death statistics.
The WHO report revealed that 10 countries alone accounted for 68 per cent of all excess deaths: Brazil, Egypt, India, Indonesia, Mexico, Peru, Russia, South Africa, Turkey and the United States.
Mr MacFeely said it was necessary to have the excess death estimates so the WHO could gain a broader measure of the impact of any crisis.
Samira Asma, WHO assistant director-general for data, analytics and delivery for impact, who co-led the calculation process, said data was the "lifeblood of public health" needed to assess and learn from what happened during the pandemic.
She called for more support for countries to improve reporting.
"Too much is unknown," she told reporters.
Not only does the data allow for countries to better prepare for crises, in India's case Rukmini said it was important for people to get a sense of "closure and validation".
"Some sort of validation to the families of people who lost their lives. There was this enormous wave of death and loss of life, much of which probably went under-counted."
"The wounds are still fresh and it's dramatic for people to recall the circumstances under which we lost people …" she added.
"I do think there needs to be some sort of conversation or reckoning around it. I do think people are owed that."
ABC/Reuters
