MUMBAI — Santosh Pandey's wife is the head of their village, population about 1,600, near the holy city of Varanasi in northern India. He sometimes answers her phone. So he's up to date on what's happening in town.
And what's happening these days, he said, is horrific.
Fifty people from his village, Ashapur, and a neighboring one, Tilamapur, have died in the past two weeks. Most died at home, gasping for breath, with fevers. But only five or six of them were officially counted as COVID-19 deaths, Pandey said.
"There's a shortage of coronavirus tests. Nobody's getting tested! So the government's numbers for our district are totally wrong," he told NPR on a crackly phone line from his village. "If you're able to get tested, results come after five days."
By that time, some of them are already gone – cremated on the banks of the nearby Ganges River.
![A health worker collects a nasal swab sample from a man to test for the coronavirus at a health center in Hyderabad. Tests are in short supply in India, where on Friday, the health ministry confirmed 386.453 infections — more than any country on any day since the pandemic began. And scientists think that's an undercount.](https://npr.brightspotcdn.com/dims4/default/e39b28c/2147483647/strip/true/crop/4859x3239+0+0/resize/880x587!/quality/90/?url=https%3A%2F%2Fmedia.npr.org%2Fassets%2Fimg%2F2021%2F04%2F30%2Fgettyimages-1232194874_custom-3849d2ad285937d23355d9ed61cc1b84fefc475f.jpg)
This village's ordeal is not atypical. Across India, there are shortages of testing kits, hospital beds, medical oxygen and antiviral drugs as a severe second wave of the pandemic crushes the health infrastructure. The country has been breaking world records daily for new cases. On Friday, India's Health Ministry confirmed 386,453 infections – more than any country on any day since the pandemic began.
Part of the reason for the huge numbers is India's size: a population of nearly 1.4 billion. The rate of known coronavirus infections per capita is still less than the United States endured at its peak. In India, there have been nearly 19 million known infections since the start of the pandemic, or roughly 1.3% of the total population – compared with more than 32 million, or almost 10% of the U.S. population.
But survivors, funeral directors and scientists say the real numbers of infections and deaths in India may be many times more than the reported figures. The sheer number of patients has all but collapsed the health system in a country that invests less on public health — just above 1% of its gross domestic product — than most of its peers. (Brazil spends more than 9% of its GDP on health; in the U.S., the figure is nearly 18%.)
![A worker carries wood to a mass cremation site for Indians who died of COVID-19. The photo was taken on the banks of Ganga River in Allahabad on April 27.](https://npr.brightspotcdn.com/dims4/default/1a89c46/2147483647/strip/true/crop/2820x2818+0+0/resize/880x879!/quality/90/?url=https%3A%2F%2Fmedia.npr.org%2Fassets%2Fimg%2F2021%2F04%2F30%2Fgettyimages-1232560954_sq-b726e3e80e21451dcde1aa9288f7a822a68bcffc.jpg)
Discrepancies in the body count
Across India, funeral pyres light up the night sky. Playgrounds and parking lots in the capital, New Delhi, have been converted into mass cremation grounds.
"The situation is very grim here. We are short of everything," said Dr. Aniket Sirohi, a municipal health official in south Delhi. "I want to say to the world, 'Help us, in any way you can.' "
Sirohi is supposed to be working on malaria prevention. But since the pandemic began, he's had the unenviable job of counting bodies. Each day, he goes to every crematorium and burial ground in his district of the capital, tallying deaths from COVID-19. Of his 11 staff members, five currently have COVID-19, he said.
"We are just doing the best we can. The morale is pretty shaken up," he said. "I have not taken a single day off."
Last year, at the height of the pandemic's first wave in India, Sirohi said he was counting about 220 COVID-19 deaths a day. When NPR spoke to him Wednesday, he counted 702 for that day. He passes those numbers up the chain of command. But the death figures the government ultimately publishes for his region have been at least 20% lower than what he's seeing on the ground, he said.
He attributed this disparity to administrative chaos. People from neighboring states flock to Delhi for medical treatment. Some die in Delhi and are cremated there but remain registered as residents somewhere else. They don't get counted anywhere, he said.
![Health workers in protective gear tend to patients in a banquet hall temporarily converted into a COVID care center in New Delhi. The photo was taken on April 28.](https://npr.brightspotcdn.com/dims4/default/a9bdd53/2147483647/strip/true/crop/4241x2802+0+0/resize/880x581!/quality/90/?url=https%3A%2F%2Fmedia.npr.org%2Fassets%2Fimg%2F2021%2F04%2F30%2Fgettyimages-1232572716_custom-d6ed68177f46da3fe1f2b785db2957d5af8fc46a.jpg)
"Somehow the numbers are not getting recorded or not shown or getting missed," Sirohi said. "India always had a poor record of maintaining these things. We have a lot of population. So there's a bit of a problem with coordination – especially in times like this, when 50% of my staff is sick."
In the western state of Gujarat, local media tracked 689 bodies that were cremated or buried under COVID-19 protocols in one day in mid-April. But just over a 10th of those deaths made it to the government's tally: The official death toll that day was 78. Such discrepancies are being reported in several states.
The "endgame" that wasn't an endgame
There is another reason why India's coronavirus numbers may be skewed: hubris. In early March, India's health minister declared that the country was in the "endgame of the COVID-19 pandemic." Daily cases had hit record lows of about 8,000 a day in early February, down from a peak of nearly 100,000 cases a day in September.
But over the winter, as cases began creeping up, some politicians didn't pay attention — or perhaps didn't believe the coronavirus could return.
Throughout April, as cases mounted, political parties held massive rallies and went door to door to ask for votes in five regions holding local elections. Prime Minister Narendra Modi presided over several rallies where thousands of supporters gathered, many without masks and with little physical distancing.
"As far as I can see, in every direction, I can only see people and nothing else," Modi boasted onstage at an April 17 rally in West Bengal. India confirmed more than 200,000 new coronavirus cases that day.
![A patient inside a banquet hall converted into a coronavirus ward in New Delhi.](https://npr.brightspotcdn.com/dims4/default/9747e5f/2147483647/strip/true/crop/4092x2725+0+0/resize/880x586!/quality/90/?url=https%3A%2F%2Fmedia.npr.org%2Fassets%2Fimg%2F2021%2F04%2F30%2Fgettyimages-1232554565_custom-8b10a9e9691a2ed6e53e8aad2262243761b7c577.jpg)
There have also been allegations that some politicians tried to suppress inconvenient news about rising case numbers.
Dr. A. Velumani runs Thyrocare, a nationwide chain of medical labs that conduct pathology tests on blood and other samples, including tests for the coronavirus. He said some of his labs have come under pressure from local officials to manipulate coronavirus tests and report fewer positive results.
"We are told, 'You shouldn't be doing more than this much,'" Velumani told local media. "In fact, in [a] good number of cases, the question is asked, 'Why [is] your laboratory reporting more positives than other local laboratories?' "
Fewer positive results mean fewer confirmed infections and fewer deaths attributed to the coronavirus. India's total pandemic deaths this week crossed the 200,000 mark. But that's still lower than the overall death tolls in the United States, Brazil and Mexico, according to data compiled by Johns Hopkins University.
There are reasons why fewer Indians might die from COVID-19. India is a very young country. Only 6% of Indians are older than 65. More than half the population is under 25. They're more likely to survive the disease.
But even considering India's demographics, scientists said the real death toll, and real number of total infections, are likely much higher than what the government is reporting.
![Family members and relatives perform last rites amid the funeral pyres for COVID-19 decedents at a New Delhi crematorium.](https://npr.brightspotcdn.com/dims4/default/15f01e3/2147483647/strip/true/crop/4409x2924+0+0/resize/880x584!/quality/90/?url=https%3A%2F%2Fmedia.npr.org%2Fassets%2Fimg%2F2021%2F04%2F30%2Fgettyimages-1232555108_custom-243d9ff69f870d0048c254c384a1189addb4b138.jpg)
Why the COVID-19 death count is likely an undercount
Mathematician Murad Banaji is a senior lecturer at Middlesex University in London. He's originally from Mumbai and has been studying the city's 2020 fatalities from COVID-19. By analyzing total excess deaths – i.e., the difference between total deaths in Mumbai one year, compared with the year before — he estimates that the number of deaths attributed to COVID-19 would have to have been undercounted by at least two-thirds to account for the higher 2020 death tally.
And that was last year when the pandemic did not strain India's health system to the extent it has now.
"My gut instinct says that right now, because you've got an overwhelmed health system, you're probably seeing an even greater undercount [of COVID-19 deaths] than you would have seen in 2020," Banaji told NPR. "My optimistic estimate then was that for every death [from COVID-19] that was recorded during the year [2020], two more were missed."
Those calculations are based on data from Mumbai, India's richest major city, where access to health care is better than elsewhere. So the number of undercounted deaths could be even higher in less well-off parts of the country — such as in Santosh Pandey's village.
It's not just deaths that are likely being undercounted. Scientists said recorded infections are even more of an underestimate. But they have a better idea of how much infections have been undercounted because they have serological data from random antibody tests that authorities conducted across large swaths of the country.
![People queue up for an inoculation at a coronavirus vaccination center in Mumbai on April 27.](https://npr.brightspotcdn.com/dims4/default/800224b/2147483647/strip/true/crop/5499x3682+0+0/resize/880x589!/quality/90/?url=https%3A%2F%2Fmedia.npr.org%2Fassets%2Fimg%2F2021%2F04%2F30%2Fgettyimages-1232553858_custom-4b9d314300567f12c825187457ae67e0c7c08dd7.jpg)
Results of a third national serological survey conducted in December and January showed that roughly a fifth of India's population had been exposed to the virus. That meant for every recorded coronavirus case, almost 30 went undetected.
But Banaji noted that capturing cases varies drastically from rural to urban areas. "You have much better detection of cases in urban areas. So for example in Delhi, about 10% of cases were being detected," he said. "Whereas in some areas of rural Bihar, considerably less than 1% of infections were being detected."
![On April 23, health workers transport a COVID-19 patient in a hospital complex in New Delhi.](https://npr.brightspotcdn.com/dims4/default/679b762/2147483647/strip/true/crop/3631x2425+0+0/resize/880x588!/quality/90/?url=https%3A%2F%2Fmedia.npr.org%2Fassets%2Fimg%2F2021%2F04%2F30%2Fgettyimages-1232489844_custom-dde5485ba4a6aa480572301e4d8a0d9b0ebda56b.jpg)
When will the wave peak?
Some scientists have been sounding the alarm about India's virus uptick since late February. Bhramar Mukherjee is one of them. She's a biostatistician at the University of Michigan who's designed models that show India's reported infections will peak in late May. She predicts India could be confirming as many as 1 million new cases a day and 4,500 daily deaths by then.
"There has been tremendous data denial. People didn't want to believe, and the government didn't want to believe, that this was actually going to explode," Mukherjee told NPR. "There was also a sense of premature celebration and exceptionalism."
Another model of India's current COVID-19 wave has been conducted by the Institute for Health Metrics and Evaluation at the University of Washington. The institute's director, Chris Murray, told NPR that India may be detecting only 3% or 4% of its daily infections.
So when India confirms about 386,000 infections, as it did Friday, that may really mean it's seeing about 10 million infections, according to the institute's modeling.
![A COVID-19 patient breathes with the help of oxygen at a tent installed along a roadside in Ghaziabad.](https://npr.brightspotcdn.com/dims4/default/298849e/2147483647/strip/true/crop/4749x3158+0+0/resize/880x585!/quality/90/?url=https%3A%2F%2Fmedia.npr.org%2Fassets%2Fimg%2F2021%2F04%2F30%2Fgettyimages-1232570338_custom-086f56b7c0bfb213e80d8aad01a085545ded606b.jpg)
"There's going to be some protection from previous infection. So you'll start to run out of people who can get infected," Murray said. "We think that infections – the actual true number of infections, which is many multiples of the officially reported cases — would be peaking probably next week."
India's deaths in this latest wave would peak around the third week of May, according to the institute's model.
"Unfortunately, given the exponential rise in cases, we may see the daily death toll rise to over 10,000 or even 12,000 a day, before we get to that peak toward the end of May," Murray said.
![Police stop vehicles at a New Delhi checkpoint aimed at enforcing a weekend lockdown imposed by the government in the wake of rising COVID-19 cases.](https://npr.brightspotcdn.com/dims4/default/9f247bf/2147483647/strip/true/crop/5141x3454+0+0/resize/880x591!/quality/90/?url=https%3A%2F%2Fmedia.npr.org%2Fassets%2Fimg%2F2021%2F04%2F30%2Fgettyimages-1232373887_custom-4a3ee29970dd7baf81aed500027ba3c97ba1c41a.jpg)
That could mean more shortages, fewer hospital beds and more tragedy on top of what India has already endured in recent weeks.
For Sirohi, the public health official in south Delhi, it would mean many more bodies to count.
"We're preparing for that, kind of anticipating it. We've acquired wood from neighboring states, to build more funeral pyres," he said. "Let's hope that doesn't happen. If it does, we'll have to face it."
Copyright 2023 NPR. To see more, visit https://www.npr.org.