Scott Cohen was on a ventilator struggling for his life with Covid-19 last April when his brothers pleaded with Plainview Hospital on Long Island to infuse him with the blood plasma of a recovered patient.

The experimental treatment was hard to get but was gaining attention at a time when doctors had little else. After an online petition drew 18,000 signatures, the hospital gave Mr. Cohen, a retired Nassau County medic, an infusion of the pale yellow stuff that some called “liquid gold.”

In those terrifying early months of the pandemic, the idea that antibody-rich plasma could save lives took on a life of its own before there was evidence that it worked. The Trump administration, buoyed by proponents at elite medical institutions, seized on plasma as a good-news story at a time when there weren’t many others. It awarded more than $ 800 million to entities involved in its collection and administration, and put Dr. Anthony S. Fauci’s face on billboards promoting the treatment.

A coalition of companies and nonprofit groups, including the Mayo Clinic, Red Cross and Microsoft, mobilized to urge donations from people who had recovered from Covid-19, enlisting celebrities like Samuel L. Jackson and Dwayne Johnson, the actor known as the Rock. Volunteers, some dressed in superhero capes, showed up to blood banks in droves.

Mr. Cohen, who later recovered, was one of them. He went on to donate his own plasma 11 times.

But by the end of the year, good evidence for convalescent plasma had not materialized, prompting many prestigious medical centers to quietly abandon it. By February, with cases and hospitalizations dropping, demand dipped below what blood banks had stockpiled. In March, the New York Blood Center called Mr. Cohen to cancel his 12th appointment. It didn’t need any more plasma.

A year ago, when Americans were dying of Covid at an alarming rate, the federal government made a big bet on plasma. No one knew if the treatment would work, but it seemed biologically plausible and safe, and there wasn’t much else to try. All told, more than 722,000 units of plasma were distributed to hospitals thanks to the federal program, which ends this month.

The government’s bet did not result in a blockbuster treatment for Covid-19, or even a decent one. But it did give the country a real-time education in the pitfalls of testing a medical treatment in the middle of an emergency. Medical science is messy and slow. And when a treatment fails, which is often, it can be difficult for its strongest proponents to let it go.

Because the government gave plasma to so many patients outside of a controlled clinical trial, it took a long time to measure its effectiveness. Eventually, studies did emerge to suggest that under the right conditions, plasma might help. But enough evidence has now accumulated to show that the country’s broad, costly plasma campaign had little effect, especially in people whose disease was advanced enough to land them in the hospital.

In interviews, three federal health officials — Dr. Stephen M. Hahn, the former commissioner of the Food and Drug Administration; Dr. Peter Marks, a top F.D.A. regulator; and Dr. H. Clifford Lane, a clinical director at the National Institutes of Health — acknowledged that the evidence for plasma was limited.

“The data are just not that strong, and it makes it makes it hard, I think, to be enthusiastic about seeing it continue to be used,” Dr. Lane said. The N.I.H. recently halted an outpatient trial of plasma because of a lack of benefit.

Plasma promotions

Doctors have used the antibodies of recovered patients as treatments for more than a century, for diseases including diphtheria, the 1918 flu and Ebola.

So when patients began falling ill with the new coronavirus last year, doctors around the world turned to the old standby.

In the United States, two hospitals — Mount Sinai in New York City and Houston Methodist in Texas — administered the first plasma units to Covid-19 patients within hours of each other on March 28.

Dr. Nicole M. Bouvier, an infectious-disease doctor who helped set up Mount Sinai’s plasma program, said the hospital had tried the experimental treatment because blood transfusions carry a relatively low risk of harm. With a new virus spreading quickly, and no approved treatments, “nature is a much better manufacturer than we are,” she said.

As Mount Sinai prepared to infuse patients with plasma, Diana Berrent, a photographer, was recovering from Covid-19 at her home in Port Washington, N.Y. Friends began sending her Mount Sinai’s call for donors.

“I had no idea what plasma was — I haven’t taken a science class since high school,” Ms. Berrent recalled. But as she researched its history in previous disease outbreaks, she became fixated on how she could help.

She formed a Facebook group of Covid-19 survivors that grew to more than 160,000 members and eventually became a health advocacy organization, Survivor Corps. She livestreamed her own donation sessions to the Facebook group, which in turn prompted more donations.

“People were flying places to go donate plasma to each other,” she said. “It was really a beautiful thing to see.”

Around the same time, Chaim Lebovits, a shoe wholesaler from Monsey, N.Y., in hard-hit Rockland County, was spreading the word about plasma within his Orthodox Jewish community. Mr. Lebovits called several rabbis he knew, and before long, thousands of Orthodox Jewish people were getting tested for coronavirus antibodies and showing up to donate. Coordinating it all was exhausting.

“April,” Mr. Lebovits recalled with a laugh, “was like 20 decades.”

Two developments that month further accelerated plasma’s use. With the help of $ 66 million in federal funding, the F.D.A. tapped the Mayo Clinic to run an expanded access program for hospitals across the country. And the government agreed to cover the administrative costs of collecting plasma, signing deals with the American Red Cross and America’s Blood Centers.

The news releases announcing those deals got none of the flashy media attention that the billion-dollar contracts for Covid-19 vaccines did when they arrived later in the summer. And the government did not disclose how much it would be investing.

That investment turned out to be significant. According to contract records, the U.S. government has paid $ 647 million to the American Red Cross and America’s Blood Centers since last April.

“The convalescent plasma program was intended to meet an urgent need for a potential therapy early in the pandemic,” a health department spokeswoman said in a statement. “When these contracts began, treatments weren’t available for hospitalized Covid-19 patients.”

As spring turned to summer, the Trump administration seized on plasma — as it had with the unproven drug hydroxychloroquine — as a promising solution. In July, the administration announced an $ 8 million advertising campaign “imploring Americans to donate their plasma and help save lives.” The blitz included promotional radio spots and billboards featuring Dr. Fauci and Dr. Hahn, the F.D.A. commissioner.

A coalition to organize the collection of plasma was beginning to take shape, connecting researchers, federal officials, activists like Ms. Berrent and Mr. Lebovits, and major corporations like Microsoft and Anthem on regular calls that have continued to this day. Nonprofit blood banks and for-profit plasma collection companies also joined the collaboration, named the Fight Is In Us.

The group also included the Mitre Corporation, a little-known nonprofit organization that had received a $ 37 million government grant to promote plasma donation around the country.

The participants sometimes had conflicting interests. While the blood banks were collecting plasma to be immediately infused in hospitalized patients, the for-profit companies needed plasma donations to develop their own blood-based treatment for Covid-19. Donations at those companies’ own centers had also dropped off after national lockdowns.

“They don’t all exactly get along,” Peter Lee, the corporate vice president of research and incubations at Microsoft, said at a virtual scientific forum in March organized by Scripps Research.

Microsoft was recruited to develop a locator tool, embedded on the group’s website, for potential donors. But the company took on a broader role “as a neutral intermediary,” Dr. Lee said.

The company also provided access to its advertising agency, which created the look and feel for the Fight Is In Us campaign, which included video testimonials from celebrities.

Lack of evidence

In August, the F.D.A. authorized plasma for emergency use under pressure from President Donald J. Trump, who had chastised federal scientists for moving too slowly.

At a news conference, Dr. Hahn, the agency’s commissioner, substantially exaggerated the data, although he later corrected his remarks following criticism from the scientific community.

In a recent interview, he said that Mr. Trump’s involvement in the plasma authorization had made the topic polarizing.

“Any discussion one could have about the science and medicine behind it didn’t happen, because it became a political issue as opposed to a medical and scientific one,” Dr. Hahn said.

The authorization did away with the Mayo Clinic system and opened access to even more hospitals. As Covid-19 cases, hospitalizations and deaths skyrocketed in the fall and winter, use of plasma did, too, according to national usage data provided by the Blood Centers of America. By January of this year, when the United States was averaging more than 130,000 hospitalizations a day, hospitals were administering 25,000 units of plasma per week.

Many community hospitals serving lower-income patients, with few other options and plasma readily available, embraced the treatment. At the Integris Health system in Oklahoma, giving patients two units of plasma became standard practice between November and January.

Dr. David Chansolme, the system’s medical director of infection prevention, acknowledged that studies of plasma had showed it was “more miss than hit,” but he said his hospitals last year lacked the resources of bigger institutions, including access to the antiviral drug remdesivir. Doctors with a flood of patients — many of them Hispanic and from rural communities — were desperate to treat them with anything they could that was safe, Dr. Chansolme said.

By the fall, accumulating evidence was showing that plasma was not the miracle that some early boosters had believed it to be. In September, the Infectious Diseases Society of America recommended that plasma not be used in hospitalized patients outside of a clinical trial. (On Wednesday, the society restricted its advice further, saying plasma should not be used at all in hospitalized patients.) In January, a highly anticipated trial in Britain was halted early because there was not strong evidence of a benefit in hospitalized patients.

In February, the F.D.A. narrowed the authorization for plasma so that it applied only to people who were early in the course of their disease or who couldn’t make their own antibodies.

Dr. Marks, the F.D.A. regulator, said that in retrospect, scientists had been too slow to adapt to those recommendations. They had known from previous disease outbreaks that plasma treatment is likely to work best when given early, and when it contained high levels of antibodies, he said.

“Somehow we didn’t really take that as seriously as perhaps we should have,” he said. “If there was a lesson in this, it’s that history actually can teach you something.”

Today, several medical centers have largely stopped giving plasma to patients. At Rush University Medical Center in Chicago, researchers found that many hospitalized patients were already producing their own antibodies, so plasma treatments would be superfluous. The Cleveland Clinic no longer routinely administers plasma because of a “lack of convincing evidence of efficacy,” according to Dr. Simon Mucha, a critical care physician.

And earlier this year, Mount Sinai stopped giving plasma to patients outside of a clinical trial. Dr. Bouvier said that she had tracked the scientific literature and that there had been a “sort of piling on” of studies that showed no benefit.

“That’s what science is — it’s a process of abandoning your old hypotheses in favor of a better hypothesis,” she said. Many initially promising drugs fail in clinical trials. “That’s just the way the cookie crumbles.”

Plasma’s future

Some scientists are calling on the F.D.A. to rescind plasma’s emergency authorization. Dr. Luciana Borio, the acting chief scientist at the agency under President Barack Obama, said that disregarding the usual scientific standards in an emergency — what she called “pandemic exceptionalism” — had drained valuable time and attention from discovering other treatments.

“Pandemic exceptionalism is something we learned from prior emergencies that leads to serious unintended consequences,” she said, referring to the ways countries leaned on inadequate studies during the Ebola outbreak. With plasma, she said, “the agency forgot lessons from past emergencies.”

While scant evidence shows that plasma will help curb the pandemic, a dedicated clutch of researchers at prominent medical institutions continue to focus on the narrow circumstances in which it might work.

Dr. Arturo Casadevall, an immunologist at Johns Hopkins University, said many of the trials had not succeeded because they tested plasma on very sick patients. “If they’re treated early, the results of the trials are all consistent,” he said.

A clinical trial in Argentina found that giving plasma early to older people reduced the progression of Covid-19. And an analysis of the Mayo Clinic program found that patients who were given plasma with a high concentration of antibodies fared better than those who did not receive the treatment. Still, in March, the N.I.H. halted a trial of plasma in people who were not yet severely ill with Covid-19 because the agency said it was unlikely to help.

With most of the medical community acknowledging plasma’s limited benefit, even the Fight Is In Us has begun to shift its focus. For months, a “clinical research” page about convalescent plasma was dominated by favorable studies and news releases, omitting major articles concluding that plasma showed little benefit.

Now, the website has been redesigned to more broadly promote not only plasma, but also testing, vaccines and other treatments like monoclonal antibodies, which are synthesized in a lab and thought to be a more potent version of plasma. Its clinical research page also includes more negative studies about plasma.

Nevertheless, the Fight Is In Us is still running Facebook ads, paid for by the federal government, telling Covid-19 survivors that “There’s a hero inside you” and “Keep up the fight.” The ads urge them to donate their plasma, even though most blood banks have stopped collecting it.

Two of plasma’s early boosters, Mr. Lebovits and Ms. Berrent, have also turned their attention to monoclonal antibodies. As he had done with plasma last spring, Mr. Lebovits helped increase acceptance of monoclonals in the Orthodox Jewish community, setting up an informational hotline, running ads in Orthodox newspapers, and creating rapid testing sites that doubled as infusion centers. Coordinating with federal officials, Mr. Lebovits has since shared his strategies with leaders in the Hispanic community in El Paso and San Diego.

And Ms. Berrent has been working with a division of the insurer UnitedHealth to match the right patients — people with underlying health conditions or who are over 65 — to that treatment.

“I’m a believer in plasma for a lot of substantive reasons, but if word came back tomorrow that jelly beans worked better, we’d be promoting jelly beans,” she said. “We are here to save lives.”