How errors and inaction sent a deadly COVID-19 variant around the world

A thick layer of dust and scatterings of dried animal excrement covered the deserted intensive care unit in Amravati, a small city nestled in a far-flung, rural district in central India. Rows of hospital beds lay empty next to sets of unplugged ventilators on a recent afternoon and only the sound of roosting pigeons punctured the ghostly silence.

The uneasy calm contrasted starkly with the chaotic scenes in early 2021 that confronted Ravi Bhushan, a 44-year-old doctor based in this cotton-growing region about 650 kilometers east of Mumbai. Toward the end of January, Amravati and its surrounding districts began to register a number of sudden and explosive coronavirus outbreaks, just as the rest of India saw a lull after a first nationwide wave of infections.

Unknown at the time, Amravati’s flare-ups were the first visible warning that the SARS-CoV-2 variant now known as delta had started along its devastating path. Within weeks, thousands of people flooded Amravati’s underfunded health care network as the city turned into Ground Zero for what would become the most confounding version of the pathogen first identified in Wuhan, China, a year earlier.

Early signs of the delta variant

Amravati was a precursor to the horrors that would grip all of India and spread globally. As January drew to a close, Bhushan was already sensing that the city of more than 600,000 residents was becoming a petri dish for a form of COVID-19 his team hadn’t treated before. Earlier, patients’ symptoms improved in under two weeks, but now they were battling the virus for “almost 20 to 25 days,” he said. “It was a nightmarish situation.”

Despite those first, ominous signs, what followed goes some ways toward explaining why two years into this pandemic, the world remains on the brink of economy-shattering shutdowns, with another new variant emerging out of vulnerable, under-vaccinated populations. But while South Africa acted swiftly in November to decode the heavily mutated omicron and publicize its existence, India’s experience perhaps better reflects the reality faced by most developing countries — and the risks they potentially pose.

India’s hampered response was characterized by months of inertia from the government of Prime Minister Narendra Modi, and a startling lack of resources, according to interviews with two dozen scientists, officials, diplomats and health workers. Many asked not to be identified because they aren’t authorized to speak to the media or were concerned about talking publicly about India’s missteps.

The actions India did — and didn’t take — as delta emerged, ultimately saddled its people and the world with a ruthlessly virulent incarnation of the coronavirus, one that challenged vaccines and containment regimes like none before it. Delta upended even the most successful pandemic strategies, snaking into countries like Australia and China with stringent “COVID zero” curbs in place and effectively closed borders. It was the most dominant form of COVID-19 for much of 2021, when more than 3.5 million people died of the virus — almost double the toll during the first year of the pandemic.

Multiple scientists interviewed by Bloomberg News said that the way India handled the early days of delta fueled its rise. The variant’s identification was delayed because the country’s laboratories were flying blind for much of 2020 and early 2021, partly because Modi’s government had restricted imports of vital genetic sequencing compounds under a nationalistic agenda to drive self sufficiency, they said. There were repeated efforts to warn the administration about the new strain in early February, the scientists said, yet India went public with details of the more transmissible variant only at the end of March.

“The charitable view is they didn’t want to sound the alarm without having more conclusive evidence,” said Ramanan Laxminarayan, director of the Center for Disease Dynamics, Economics & Policy, a research institute with offices in New Delhi and Washington that has helped conduct some of India’s largest COVID-19 studies. “The more skeptical view is that everything about COVID has been politically managed and therefore it’s a question of whether they didn’t want to alarm people about something happening in India for fear of stigma — almost every country is trying to play this game.”

While the fallout from delta has been devastating, many of its lessons are yet to be addressed. South Africa’s vigilance makes it an outlier, with gaping virus surveillance holes across the world, particularly in countries that have already limited means to purchase and distribute vaccines, let alone the luxury of expensive sequencing programs.

More than 80% of the 6.5 million SARS-CoV-2 genomes that have been decoded and then uploaded to GISAID, the international database that tracks changes in the virus, have come from Europe and North America. Given the way southern Africa was treated when news of omicron emerged — with travel curbs levied from the U.S. to Japan — there may also be limited political appetite to remedy that disparity.

Tracking for omicron versus delta

Genome sequencing, the process used to decode the coronavirus and learn more about the rhythm of mutations, stands at the heart of the world’s attempts to defend itself from new waves. What happened with delta shows how the absence of a strong virus-monitoring infrastructure, especially for sequencing, can turn crowded cities into spreading grounds for dangerous new variants — with the potential to prolong the pandemic.

Even though omicron is quickly becoming the more dominant form of COVID-19 in the U.S. and elsewhere, quick action has bought time for scientists to decode the extent of its transmissibility and severity. South Africa identified and broadcast details of the new variant just weeks after seeing a spike in cases in one province.

By contrast, for much of 2020, India’s efforts tracking the virus were sparse, meaning the exact origin of delta still remains murky. To date, the country has only sequenced and shared 0.3% of its total official infections to the GISAID database.

India has been held back by the fact that only a handful of government laboratories and states were making consistent efforts in the first year of the pandemic to map the virus, even as millions were being infected in the country’s first wave, according to people familiar with the matter.

Bhramar Mukherjee, an epidemiologist and biostatistics chair at the University of Michigan’s School of Public Health, said India’s sequencing efforts were hurt by “bureaucracy, politics and a sense of exceptionalism that we have conquered COVID and there is no need to worry about variants.”

“The need to share data and samples is so key,” she said. “When South Africa started collaborating and sharing with the rest of the world, progress also increased like a process of contagion: exponentially. India is always protective of its own data.”

Inside India’s scientific agencies, a lack of institutional dynamism, along with a culture of subservience to Modi’s government — highly sensitive to commentary on its handling of the virus — had taken hold, said one former official. That meant critical questions weren’t being aired by experts out of fear they’d derail their careers, the person said. In many cases, India’s health ministry simply wasn’t listening to or making decisions based on advice coming from those expert bodies, according to this official.

Attempts to ramp up sequencing in India were also critically curtailed by an inadvertent ban in May 2020 on the import of reagents, the chemical needed to fuel sequencer machines. The “Make in India” campaign, Modi’s drive to ensure the country is less reliant on places like China, meant publicly financed labs weren’t able to import items worth less than 2 billion rupees ($26.5 million) for months. India mostly uses sequencers manufactured by California-based Illumina Inc. and the U.K.’s Oxford Nanopore Technologies PLC, which run on patented reagents that can’t be substituted locally.

Scientists in India and abroad now provide varying dates for when delta began circulating there. Samples retrospectively added to GISAID show at least one delta-linked lineage in India as far back as September last year, while the World Health Organization places its first discovery there in October.

Current and former Indian government scientists say there are often errors when manually uploading information to the database and those datelines are likely to be wrong. December 2020 is when delta was initially sequenced in India, they say. Certainly, the first person to decode the mutations wouldn’t have known its full enormity at the time since not all changes in a virus are significant. Only when you begin to see spiraling outbreaks marked by similar characteristics do you realize that a variant of concern is at play, they said. But Amravati offered the clues needed to make that connection as early as January 2021.

India isn’t the only nation where scientists have struggled to convince governments about the merits of decoding the structure of the virus. Even in the U.K., which has a world-leading coronavirus sequencing drive that has recorded 13% of all cases, there was early criticism that it was essentially an expensive academic exercise of limited real-world value.

The U.S. also reported omicron cases much later than other countries, even though officials said the new variant was likely already in the country, a sign of gaps in its sequencing systems.

Anticipating COVID-19’s evolution

Some Indian scientists were already convinced by the summer of last year that the virus would morph in the country in a way that would have grave implications. But after India’s initial wave of cases subsided in October 2020, defying predictions for an explosion, officials in Modi’s administration were less inclined to listen to scientific concerns, the experts who spoke to Bloomberg said.

The emergence of the alpha variant in the U.K. in mid-December 2020 reinvigorated the urgency around sequencing for some in India’s scientific community. A long-standing request by one government lab for Illumina reagents was soon working its way through several layers of Indian bureaucracy. The government had recently given the green light to launch the Indian SARS-CoV-2 Genomics Consortium (INSACOG), a grouping of 10 state-funded laboratories set up to help plug the large holes in India’s genomic network, and the nation’s scientists knew they had to act fast.

Calls on Dec. 24, 2020 — made with apologies for the festive disturbance — eventually helped arrange shipments of badly needed reagents and equipment from the U.S. and Singapore. While the import rules that stymied access to lab materials were eased in January, a shortage of flights because of global travel curbs provided the next hurdle. The scientists managed to get some of the items they needed, but it would still be weeks, and in some cases months, before many of the INSACOG labs were fully equipped.

Around the same time, many Indians were convinced their country had already borne the brunt of the disease. That was also true in remote locales such as Amravati, where residents were seeking to rebuild their lives after the national lockdown that Modi — who has styled his premiership on often authoritarian, swift and singular action — had imposed with just a few hours notice in late March 2020.

But early in 2021, Prashant Thakare, an associate professor of biotechnology at the Sant Gadge Baba Amravati University, was deeply worried. He was seeing a sudden spike of infections at the testing station housed on campus, which often plays host to wandering leopards from the nearby forests. “Complete families were testing positive,” said Thakare, who said he rushed to notify the district’s top health officials. “This was not the case in the first wave, so that took us by surprise.”

The surge of the delta variant

The government of Maharashtra, the state where Amravati is located, was responsive, Thakare said, and sought to get a handle on the situation. But even when you have the required equipment — which many labs still didn’t — sequencing in India was a time-consuming process. Given the country’s vastness and patchy infrastructure, some virus swabs could be 10 days old by the time they reached scientists, who would then often wait weeks for enough samples to warrant utilizing their expensive-to-operate sequencing machines. The quality of many of the specimens was often also low, sometimes handled by inexpert, indifferent or harried front-line health workers.

At this point, INSACOG, the sequencing consortium, was also struggling to get off the ground. Though formed at the end of 2020, many of the participating labs were months later still undergoing training and trying to source equipment. Facilities under the Indian Council of Medical Research — one of the bodies most closely advising the government on COVID-19 — were refusing to share genomic data with other INSACOG partners, according to a person familiar with the matter.

An early promise to INSACOG of 1.15 billion rupees from the Prime Minister’s Citizen Assistance and Relief in Emergency Situations Fund — set up by Modi in March 2020 to provide emergency grants — also failed to materialize, leaving the Department of Biotechnology to stump up the money, only 700 million rupees of which was released by the end of March 2021, the person added.

Searching for answers in a crisis

By early February this year, with hospitals already overflowing, health officials in Amravati knew they needed answers fast and rushed about a dozen positive swabs from the city and neighboring districts to the city of Pune. Roughly three hours drive inland from Mumbai, Pune is home to a number of research institutes studying the virus, as well as the Serum Institute of India, the giant vaccine-maker key to India’s COVID-19 immunization drive. The samples needing urgent decoding were delivered to the BJ Medical College, an expansive white building adjoined to the Sassoon General Hospital, a storied state facility that had once given Mahatma Gandhi an emergency appendectomy.

There, Rajesh Karyakarte took charge. The enthusiastic and jovial head of the college’s microbiology department has scientific endeavor running through his blood: A parasite discovered by his scientist father was given the family name. At the start of the pandemic, Karyakarte had convinced the college to purchase a MinION portable sequencer from Oxford Nanopore, a device weighing nearly 500 grams that’s around the size of a small chocolate bar.

The sequencer finally arrived on Feb. 1, 2021, and the following day Karyakarte eagerly agreed to analyze the samples from Amravati after Maharashtra’s director of medical education and research raised concerns about the soaring infection rate there on a weekly video call. Within two days, a dozen specimens from the region were in his care. There was just one hitch. The sequencer had yet to undergo the mandatory in-person installation by company representatives based in India’s southern tech hub of Bengaluru, who were barred from traveling during the pandemic.

“We were in a fix,” Karyakarte said.

His team scrambled to come up with a solution. One of Karyakarte’s doctoral students volunteered to drive the samples about 14 hours to Bengaluru, where they could be trained in person to use the Oxford Nanopore sequencer. But they soon realized the dry-ice packaging essential for keeping the samples stable would likely melt before they got there. After some detailed negotiations, Air India agreed to let them transport the virus samples in their cargo hold.

On the afternoon of Feb. 8, three of Karyakarte’s students flew down to Bengaluru and over the next three days set about unlocking Amravati’s secrets.

When a state investigative team arrived in Amravati in the middle of February, they were shocked to find entire communities were falling sick. One person on that trip said they shared those concerns with India’s health secretary, the heads of national agencies, and anyone they thought had the clout to make policy changes, warning them the country was heading for very serious trouble and that the virus appeared to have mutated. Officials in New Delhi said that the information would be verified and appropriate action taken, but in the face of growing evidence of a public health emergency the response was deeply frustrating, the person said.

When the raw sequenced data arrived back in Pune from Bengaluru, Karyakarte’s team focused on making sense of it. The Amravati samples had two curious mutations on the spike protein — the key the virus uses to unlock the doorway into our body’s cells. Karyakarte presented his concerns in detail to state officials on Feb. 18.

“Those mutations were immune escape mutations,” he said. “I red flagged that.”

Revealing delta to the world

The next day, Karyakarte’s report was sent to the nearby National Center for Cell Science and the National Institute of Virology in Pune for his analysis to be cross-checked. Jointly, they found B.1.617, the virus lineage that would break into three and spawn delta. Provided with evidence of the mutations, along with what was by then a soaring case rate, the Maharashtra government locked down Amravati on Feb. 22.

Thakare, whose PCR lab was among the first to pick up on the infection surge, says he was told by state authorities not to disclose Karyakarte’s findings, and that other health officials in the district were kept in the dark about the new variant. Calls and emails from Bloomberg to Maharashtra’s health minister, Rajesh Tope, weren’t answered.

Around March 10, INSACOG filed an internal report to the National Center for Disease Control — the agency heading the consortium — confirming that a new variant was spreading and was a matter of high concern, according to people familiar with the matter. There was no public pronouncement until a March 17 meeting with the chief ministers of India’s states, when Modi acknowledged the rising caseload in Maharashtra and warned of a growing national outbreak, along with the need to identify coronavirus mutations and their effects. It was only on March 24, however, that India’s health ministry confirmed the presence of the double mutation. It conferred “immune escape and increased infectivity” and was showing up in as many as 20% of samples, the ministry said, but it hadn’t “been detected in numbers sufficient” to explain the rapid rise of infections.

A former official said they were flummoxed at why it took so long to make knowledge of the variant public. Two weeks in a pandemic is a lifetime, the person said. But other scientists claim it wasn’t clear then that the strain was indeed the cause of what was fast becoming a nationwide jump in infections: By the end of March, India was seeing more than 60,000 cases a day, from less than 15,000 a month earlier.

“There was a little time lag before we associated the surge uniquely with the delta variant,” said Priya Abraham, director of the state-funded National Institute of Virology. “People didn’t really associate at that time this ‘dual variant’ with all that was happening in the country until it became apparent that this variant was now riding the crest and being detected in many other states.”

N.K. Arora, the co-chair of INSACOG and a member of India’s COVID-19 taskforce, said in November that the country was still learning how to manage the pandemic in early 2021 and in hindsight it was very easy to ask why there was a delay. Officials from Modi’s office, the health ministry, the department for biotechnology, the National Center for Disease Control, the Indian Council of Medical Research and Niti Aayog — a government think tank and policy adviser — didn’t respond to multiple requests for interviews or comment on a detailed list of questions.

But with little action taken, the conditions were ideal for COVID-19 to spread. A number of state elections that Modi and his party vigorously contested were allowed to proceed through March and April, with huge crowds drawn to rallies and polling stations. The Kumbh Mela, the largest religious gathering in the world, saw millions of Hindu pilgrims gathering cheek-by-jowl along the banks of the holy river Ganges in April. The result was like upgrading from a pistol to a machine gun, said William Haseltine, the founder of Human Genome Sciences Inc. who now chairs Access Health International, a New York-based think tank with offices in India. Spray “a thousand bullets into a crowd, the chances you’re going to get hit is a lot higher,” he said.

Debating pandemic policy

India’s government, like many across the world, was loathe to impose fresh pandemic restrictions after the economic damage of the 2020 lockdown. Modi left those decisions to India’s state governments, which eventually began to implement a confusing patchwork of curbs. New Delhi was also fixated on the naming of B.1.617 — bristling at its initial description as the “Indian variant” in an echo of the government’s anger almost a decade ago when antibiotic-resistant superbugs named after the capital threatened to dent the country’s image and its lucrative medical tourism industry.

Case rates in India had shot up so much by mid-April that the tide of infections became impossible to ignore. On April 12, Indian scientists presented their findings on B.1.617 to the WHO. At a media briefing toward the end of that week, the agency designated it a “variant of interest,” though Maria Van Kerkhove, the WHO’s COVID-19 technical lead, called it “concerning.” That same day, Indian health experts continued to downplay the threat.

“This variant was present from January onwards but the surge in cases that we are getting in this wave is just from the beginning weeks of March,” Aparna Mukherjee, a senior scientist at the ICMR, told Bloomberg Television on April 16. “We are not sure if it’s this variant causing the surge.”

By then, that debate was largely academic — the strain had exploded across India, and would soon spread across the world. Kamlendra Singh, an assistant professor at the University of Missouri, arrived in north India to visit family in late March. Just weeks later, he remembers “it was going really nuts in India.”

The biologist was well aware of the dangers the virus presented, having published a number of research papers on the disease. For the rest of his trip he took precautions and limited his interactions, even though he had been double vaccinated with Pfizer Inc.’s shot. After testing negative, Singh boarded a flight back to the U.S. at the end of April.

In the air, he felt a weird tiredness set in. Landing in Chicago, the professor declared his symptoms to immigration, but with his negative test results in hand they waved him through to a connecting flight to St. Louis. On the drive home to Colombia, he started to feel progressively worse. The next day Singh went to get swabbed again. His test came back positive. “The entire plane was full,” he remembers. “Who knows how many were infected in there?”

Through countless trips like Singh’s, B.1.617.2 — the numeral “2” is used to distinguish delta from its parent — started seeding itself across the world. The U.K.’s extensive sequencing network soon picked up the variant on its soil in April and within weeks scientists there estimated that it was possibly at least 50% more transmissible than the then predominant alpha strain.

Still, Boris Johnson’s government was slow to erect travel curbs against India as it eyed a trade deal with the country, even though by early April it had red-listed neighboring Bangladesh and Pakistan, which were reporting fewer infections. On April 14, the prime minister’s office announced an India trade trip had been scaled back from four days to just one. Downing Street then canned the visit five days later and added India to its red list of banned travelers.

The new measures, though, didn’t take effect until April 23 and Britain soon became a hotbed for the new strain, providing it with another springboard to spread. The variant added urgency to the U.K.’s vaccination drive and pushed back the country’s economic reopening plans as European countries imposed their own new barriers. It also went on to spawn devastating outbreaks elsewhere, overwhelming hospitals from Kentucky to the Philippines.

By mid-May, the strain had been re-labeled a “variant of concern” by the WHO, which by the end of that month had put in place a new naming system based on the Greek alphabet — christening B.1.617.2 as delta.

For India, the trauma of the delta-fueled wave was absolute. Many lost family and friends, or at least knew someone who had. Tragic scenes from hospitals, like those in Amravati, were beamed around the world, laying bare the failings of a health system that struggled to even provide basic care such as oxygen and hospital beds.

It was a wake-up call that’s seen the country modestly bolster its sequencing efforts, though the original plan to decode 5% of all infections was abandoned as cases peaked at more than 400,000 a day in May. Twenty-eight new labs have since been added to INSACOG, a body that state scientists hope will become a permanent fixture beyond the pandemic.

One facility added to that roster was the BJ Medical College in Pune. A visit in mid-October to Karyakarte’s large office with blue, water-stained walls found him surrounded by about a dozen doctoral students. Freshly arrived that day was a new and larger sequencer. It was fortuitous timing: While omicron was still weeks away from making headlines, Karyakarte, now Maharashtra’s coordinator for genome sequencing, was worried about India’s inoculation levels, which at 42% double vaccinated is well below similar-sized countries like China and Brazil, according to Bloomberg’s vaccine tracker.

“Corona has not gone anywhere,” he said as the students went off to set up the new sequencer. “The tail is going to be very long.”

Back in Amravati, the district’s chief surgeon, Shyamsundar Nikam — who had a near-lethal brush with delta himself — is similarly cautious. While just a few months off retirement, he pushed for massive new oxygen canisters to be built around Amravati’s hospitals. “We have to keep it ready because of that omicron,” said Nikam, who sent pictures in mid-December of the now spotless ICU unit, scrubbed clean of dust, grime and animal droppings.

Recovering from catastrophe

India’s government is also trying to hasten the immunization drive, though it will start inoculating those under 18 and offer health care workers a booster dose only next month. At a vaccine center earlier this month in Amravati, a steady crowd queued to get their shots. Around the world, while omicron — with its some 50 mutations — extends its sweep, some countries seem better placed to handle what lies ahead. “We were not caught with our pants down,” Salim Abdool Karim, an epidemiologist who has advised the government in South Africa, said last month. “We expected and we were ready for a new variant.”

For now, the hope is that, unlike delta, omicron turns out to be a milder version of the virus, one that seeks to replicate as much as possible without killing off its hosts. But with so many people remaining unvaccinated, especially in sub-Saharan Africa, the risk of more strains emerging remains, said Brian Wahl, an epidemiologist at the Johns Hopkins Bloomberg School of Public Health. Less than 9% of Africa’s population is fully inoculated.

“In order to stop the emergence of a new potentially devastating variant we need to make sure that the world is protected,” Wahl said. “That was the lesson from delta. Have we acted on that? That’s a different question and why we’ve seen the emergence of omicron.”