This story is one in our six-part series The Pandemic Playbook. Explore all the stories here.
Peter Horby couldn’t believe his eyes.
He’d just gotten his first glimpse of new data from the United Kingdom’s Recovery Trial, an experiment enrolling tens of thousands of patients at dozens of hospitals for clinical trials investigating Covid-19 therapies.
Horby and his Oxford University colleague Martin Landray had dreamed up the trial program in March 2020. Nearly three months later, more than 35,000 people were dead in the UK (and more than 100,000 in the US), and Horby and Landray were still hunting for the world’s first lifesaving treatment for the novel coronavirus.
Now Horby was looking at the results of a Recovery Trial project testing dexamethasone, a cheap, widely available steroid. It found dexamethasone had lowered the likelihood of death for Covid-19 patients who required oxygen or a ventilator while in the hospital.
He called Landray. His summary was pithy.
“Holy shit, it works.”
On June 16, after verifying the data, the Recovery Trial announced the news: A drug had finally been shown to reduce Covid-19’s mortality. Dexamethasone became part of the standard of care across the world. A UK government estimate published in March 2021 concluded that the drug’s use had so far saved 22,000 lives in the UK and an estimated 1 million lives worldwide.
As the coronavirus pandemic exploded across the world, medical science was starting from scratch. While the long-range goal was an effective vaccine, millions could — and would — die before a shot was approved and widely distributed. The best bet for preventing as many deaths as possible was to find existing drugs that could treat Covid-19.
But doing this would require enormous capacity: Researchers had to test a lot of different therapies simultaneously, which would mean recruiting thousands of patients to participate. In the US, the New York Times’s Carl Zimmer wrote in January 2021: “many trials for Covid antivirals were doomed from the start — too small and poorly designed to provide useful data.”
The United Kingdom’s Recovery Trial was the opposite: massive and simple. It has proven to be the most effective program in the world for delivering desperately needed research results. Pharmaceutical companies sought out partnerships with the Recovery Trial because they see it as the most reliable way to determine whether their drugs can help stop the global outbreak. Since June 2020, when Horby and Landray saw the first results on dexamethasone, the trial has since identified another drug, tocilizumab, that improves mortality rates. It has also done the valuable service of demonstrating that some once-promising treatment candidates don’t work.
In this series, the Pandemic Playbook, Vox is exploring the successes — and setbacks — of countries around the world in combating Covid-19. In many important ways, Britain struggled; it has, per capita, suffered among the most cases and deaths of any country in the world. But the Recovery Trial is a notable exception. It’s made vital contributions to medicine’s understanding of Covid-19 that saved lives not just in the United Kingdom but around the world.
The US health care system, for all its flaws, is often touted for encouraging innovation. But the Recovery Trial was made possible by the UK’s unified, government-run health care system. The National Health Service runs most hospitals in the country, and their staff, including the research staff, are all government employees. British patients’ medical records are all shared in one data system.
This made setting up the world’s biggest clinical trial in the middle of a pandemic straightforward. To hear the doctors involved tell it, it was as simple as the country’s chief medical officer sending a letter to the head of every hospital, asking them to take part. The grand experiment was soon underway.
“We were able to get up and going within days. That was because of the NHS,” says Duncan Browne, who is running the Recovery Trial program at a hospital in Cornwall. “Whereas if we had a fragmented approach, I think we’d still be arguing about it.”
The ideas behind the Recovery Trial can be traced back to a very different program in the 1980s
The road to the Recovery Trial started in the 1980s, when a group of Oxford scholars was dissatisfied with the lack of treatments for heart attacks. They imagined a trial that could test different interventions — a massive trial, perhaps as many as 10,000 to 15,000 patients. For such a big trial to work, it had to be simple: Nurses and doctors would need to be able to try out the treatments the researchers were testing as part of their normal care routine.
The system they set up began when a patient was admitted to the hospital. The research team would tell the medical staff which treatment pack to use for the patient, a pack containing either a placebo or a treatment. (Nobody, including the researchers, knew which.) The hospital staff had to report whether the patient lived or died, plus any notable events, like a blood clot or a stroke.
Ultimately, that research, the International Studies of Infarct Survival, enrolled more than 140,000 patients and identified several treatments that significantly cut the number of heart attack deaths, including a combination of an anti-clotting enzyme and aspirin — a groundbreaking finding, and a treatment regimen still in use today. They became famous in the medical field, attracting younger medical researchers like Landray, who came to Oxford in 2000 to work with the team.
But nobody had set up a very large clinical trial program in the middle of an emergency before. Previous efforts had failed.
Horby, who was already studying infectious diseases before the pandemic, had given lectures about the missed opportunities for research during the SARS crisis: “all the studies they could have done but didn’t.” He had tried to get some clinical trials started during the H1N1 flu pandemic in 2009, but he says it had taken researchers three weeks to draft the proposals, which then spent two more weeks in front of an ethics review board before getting a green light. The trials that made it through the process enrolled just a few hundred patients, not nearly enough to yield robust results.
“It was appalling. Hardly any trials started,” Horby says. “It just was a failed model for epidemic infections.”
Landray, meanwhile, had spent his life before Covid-19 trying to figure out how to best run big — really big — clinical trials. But he’d never given infectious disease outbreaks much thought until the coronavirus pandemic began.
On February 28, when there were still just 20 confirmed Covid-19 cases in the UK, Landray emailed Jeremy Farrar, who leads the Wellcome Trust, one of the top medical research charities in the country. It seemed clear to Landray that the SARS-CoV-2 virus would spread fast, and clinical trials for possible treatments would be needed. Vaccine research had barely begun, and in the short term, doctors would need to find something else to save lives — likely a treatment for an existing condition that could also work against Covid-19.
Farrar suggested that he contact Peter Horby, who had secured financial support for a multi-drug trial in China. But by the time the financing was approved, the Chinese outbreak had slowed down — while cases were taking off in the UK.
When Horby and Landray sat down for their first face-to-face meeting in Oxford on March 5, they realized they each had one half of the equation. Horby had the protocols to set up a multi-drug investigation during the pandemic; Landray had been thinking about how to establish such a massive trial in the UK.
“He’d never done a clinical trial over about 200 patients,” Landray says. “I’d never, ever worked on infectious diseases.”
They had a proposal drawn up by March 10. With Farrar’s support, they pitched it to England’s chief medical officer, Chris Whitty. Whitty bought it.
He and his peers in Scotland, Wales, and Northern Ireland sent out a missive to the NHS hospitals, urging them to take part in this public project. More than 175 hospitals across the country eventually agreed to participate.
“Back in March, we didn’t really know what was going on. It was more scary than it is now,” Browne says. “To feel that you were doing something was, I think, a really positive thing.”
Within two weeks of Horby and Landray’s first meeting, more than 1,000 patients had already enrolled in the Recovery Trial. Now all they had to do was wait.
Once it got started, the Recovery Trial successfully found effective Covid-19 treatments
Browne, the lead investigator at the Cornwall hospital, and Fiona Hammonds, who serves as the lead research nurse there, had been studying diabetes before the Covid-19 pandemic. Neither of them had worked on infectious diseases before.
But that lack of experience didn’t matter, Browne says. “It’s really more about the process of doing research.”
Horby and Landray had designed the Recovery Trial, like the heart attack trials before it, to be as easy to administer as possible.
Recruitment started as soon as a patient walked through the hospital doors. UK emergency rooms were stocked with pamphlets and posters advertising the Recovery Trial. Within three or four hours of testing positive for Covid-19 and being admitted, an eligible patient would be approached by a member of the research team and asked if they would like to take part in the trial. Many of them wanted to enroll, Browne and Hammonds found. More than 35,000 patients have been enrolled in the Recovery Trial over the past year.
Once a patient agreed to participate, they signed a simplified consent form — one page instead the more typical five or six — shorn of a lot of the fine print. The research team would then enter the patient into the Recovery Trial’s computer algorithm, which randomly assigned them to one of the trials. The NHS’s electronic system then generated a prescription for the experimental drug for half the patients; the other half served as the control group. Dexamethasone, or one of the other drugs tested in the trial, became just another name on a patient’s drug chart, given by the nurses in the normal course of care.
“It became no more time for the nurses and doctors in the ward than if it had been a conventional part of their treatment,” Browne says. The patients who were given the treatment in the trial were then compared to the patients who received the regular standard of care.
Nurses could be trained to run the trial in about 20 minutes. They logged the patient’s discharge date as well as side effects or adverse events. The information was automatically uploaded to an online NHS database, where an independent panel monitored the data.
A group of experts formally recommended which treatments should be included in the trial. Some of Recovery’s candidates were obvious because they were already being used for Covid-19 care; hydroxychloroquine, hyped by then-US President Donald Trump as a possible cure for Covid-19, was one such candidate.
Others, like dexamethasone, were more educated guesses. There was limited research available on dexamethasone and viral infections. Some of it showed the drug could be helpful, but there were also indications it could be dangerous at high doses. Nobody knew for sure if it would work. Prominent experts warned against testing dexamethasone, Horby and Landray say, because the potential adverse effects made it unethical.
The investigators saw it differently. There was already an existing proposal to set up a steroid trial whenever the next flu pandemic happened. They thought the uncertainty was a good reason to conduct a trial.
“There was a lot of controversy about it, which probably tells you it’s a good drug to test,” Horby says. “Opinion was very divided, and this was the only way to un-divide the opinion.”
In one year, the trial has examined nine repurposed drugs, a convalescent plasma, and a newly developed antibody cocktail.
And it has found that two of those drugs reduce mortality for seriously ill Covid-19 patients: dexamethasone and, more recently, tocilizumab.
Paul Buckler, 46, was one of the 4,000-some patients enrolled in the tocilizumab trial. He lives near Southampton, almost two hours southwest of London. He says he’d been careful during the pandemic — his dad has chronic obstructive pulmonary disease — but he still found himself infected with Covid-19 in early November.
He first noticed he couldn’t smell his hand sanitizer. Tipped off that something was wrong, Buckler got a coronavirus test — and it was positive. At first, he didn’t feel particularly bad, but after a few days, he started coughing and lost his appetite. He checked his blood oxygen levels with a home monitor and his numbers looked low. He called his primary care doctor, who told him to get himself to a hospital.
“I knew I wasn’t right,” he says.
The morning after his first night in the hospital, a nurse and a Recovery staff member came to Buckler’s room and pitched him on participating in the trial. They reassured him they knew these drugs well, including what side effects to expect. They gave him some material to read, though he admits he didn’t examine the fine print. He said he’d do it, and by that afternoon, the hospital staff told him he had been selected for the tocilizumab trial.
He never knew if he got the placebo or not. But he did start feeling better in a matter of days.
“I didn’t even think about it. When you feel ill, if someone says, ‘We’ve got these drugs that might help you get better,’ you think, ‘Okay,’” Buckler says. “I did do my bit a little bit, I guess. Everybody’s doing a bit. That’s what Covid has really shown: If everybody does their bit, it does help.”
In February 2021, the results were published: Tocilizumab, largely when taken in tandem with dexamethasone, reduced mortality for patients who were receiving oxygen (like Buckler) or placed on a mechanical ventilator. It also helped patients who developed severe symptoms be discharged from the hospital more quickly.
The finding was another victory for the Recovery Trial’s premise of going big. Earlier, smaller trials had failed to show any benefit in using tocilizumab. Now doctors could put it in their care regimen with confidence, right away.
“We publish the results at lunchtime,” Landray says, “and it’s practiced by tea time.”
Critics say the Recovery Trial is an imperfect program — but maybe still the best option in a crisis
The UK’s research achievements during the pandemic extend beyond the Recovery Trial.
Two other trial programs — one for patients in primary care, another for only the critically ill — ran alongside it. Browne’s hospital participated in the Siren study, which tracked antibodies in hospital workers. British scientists also lead the world in sequencing coronavirus genomes, work that was critical in identifying new variants. And while the US has played an important role in developing and producing Covid-19 vaccines, the UK hasn’t been too shabby, either: The country has vaccinated about the same share of its population as the US, with almost half of those doses being the AstraZeneca vaccine designed by Oxford scientists.
But the Recovery Trial especially helped set British science apart during the pandemic. Even its critics preface their critiques with praise for its scope and extraordinary efficiency amid a public health emergency, deploying descriptors like “brilliant” and “absolutely stellar.”
“Covid was new and we had no idea how to treat it,” says Tobias Kurth, a Berlin-based epidemiologist who focuses on study design. “The approach was really good.”
But precision is sacrificed when you design a study as simple as the Recovery Trial. Big numbers are good at answering basic questions — did the patient live or die? — but they don’t provide the same level of detail or certainty that researchers are accustomed to.
The investigators know they are making a trade-off by keeping it simple. Horby spoke at the WHO about the tocilizumab results, and he remembers somebody asking him whether the drug would be clinically appropriate for a diabetic in their 70s.
“I can’t answer the question,” he says. “We didn’t look at that specific subgroup.”
And the Recovery Trial has at times been labeled with the four words that have stained other scientific findings during the Covid-19 pandemic: “science by press release.” Announcing initial study results without a full data set to support them could erode the trust among experts whose acceptance the researchers need to achieve widespread adoption of their proven treatments.
For Kurth, from a methodological perspective, the Recovery Trial’s core challenge is how it conveys certainty. To give an overwrought example: Does dexamethasone really “work,” as Horby said in disbelief during his phone call with Landray? Or are there “indications that it works”? For people who think hard about how to design scientifically viable studies, there is a big difference between a signal — some kind of association that may indicate, but does not prove, a cause and effect — and a real result, a demonstrable causal link. Kurth says it’s still hard to know which one we’re getting from the Recovery Trial.
The push for speed might have another downside, Kurth said. Recovery Trial data is published when an independent panel of experts, which monitors the incoming data, decides it has seen enough.
Kurth’s concern is that they might be making the call too soon. What if data collection had continued, and the effect that the study appeared to detect when it was stopped actually disappeared? Tocilizumab, for example, saves one person who otherwise would have died for every 25 patients who take it. It’s at least conceivable that, with even more data, that link could evaporate.
This was another methodological risk Horby and Landray knew they were taking. But they feel a responsibility to publish clinically relevant information they believe could save lives. Even a small effect means thousands of deaths averted, given the number of people who ended up in the hospital with severe Covid-19.
“My own view has been: We had confidence in our data,” Horby says. “Are we going to sit on it when there are hundreds of thousands of people in the hospital and we know the drug is going to work?”
The UK outpaced the US in researching Covid-19 therapies because it has a unified health system
The United Kingdom had one big advantage for researchers looking to set up a massive clinical trial in no time flat: the National Health Service.
The NHS made it easier for the Recovery Trial to launch with the urgency required by the moment. Derek Lowe, who writes about drug discovery, told me last summer that the US could have conceivably set up a similar program.
But it didn’t. Instead, clinical trials in the US were forced to try to coordinate among disconnected hospital systems. The NHS is a single system of roughly 1,250 hospitals; the largest hospital system in the US includes fewer than 200 facilities and most are much smaller than that. They struggled to achieve the same scale as the Recovery Trial — and therefore were slow to deliver results. As a result, even American companies developing new therapies in the US ended up looking to the Recovery Trial when it was time for large-scale tests.
One was Regeneron, a US-based biotech firm that developed a cocktail of Covid-19 antibodies. They initially ran a small clinical trial in cooperation with America’s biodefense research agency, BARDA, but the company realized it would need a much bigger sample size to get reliable results on whether their drugs reduced deaths.
So when they were ready to start wider testing in September, they decided to partner with the Recovery Trial. The US trial had enrolled just a few hundred patients; the UK trial would eventually include data from more than 9,000 people.
“In order to see really meaningful results, you really needed a very large sample size,” Leah Lipsich, vice president at Regeneron, says. “Working with Recovery, a big-basket trial, was the way to go.”
A reduction in Covid-19 hospitalizations in the UK, likely driven by the country’s successful vaccine rollout, has slowed down enrollment in the trial. But Regeneron expects results in the next few months.
Lowe and Lipsich both thought the NHS had probably helped the UK succeed in this difficult task.
“The NHS is an enormous network, and all of those hospitals are linked together,” Lipsich said, describing a single system that was seeing, in effect, every Covid-19 patient in the country. “I don’t think you can say that about any place in the US.”
The UK pulled together in a moment of crisis, when the prognosis felt grim. Hammonds, the nurse in Cornwall, found comfort in the first dexamethasone result. She knew she could now tell patients when recruiting them to join the Recovery Trial: We’ve already found one thing that worked, and it will help you.
“We know what the best treatment is, and you’re receiving it,” she says. “Would you like to receive one or two more?”
Alicia Canter is an independent photographer based in London.
This project was supported by the Commonwealth Fund, a national private foundation based in New York City that supports independent research on health care issues and makes grants to improve health care practice and policy.