The good news that an inexpensive generic drug cut deaths by a third in severely ill COVID-19 patients was cheered — and criticized — by physicians in the Philadelphia region and around the world on Tuesday.

They were delighted to hear such encouraging results from a rigorous British study of a corticosteroid called dexamethasone. But they also were disturbed that those results came out piecemeal in a self-congratulatory news release, rather than a journal article reviewed by outside experts before publication. During the pandemic, companies and researchers have been racing to share results without adequate scientific vetting, a trend that has contributed to backpedaling and confusion about potential therapeutics.

“I would call the [British] results intriguing,” said George Anesi, a pulmonary and critical care specialist at the Hospital of the University of Pennsylvania. “There has been a lot of rapid publication, some of which has required correction or retraction. We’d want to see the paper before making changes in practice.”

Surgeon Atul Gawande, the public health researcher and author, tweeted, “It will be great news if dexamethasone, a cheap steroid, really does cut deaths by ⅓ in ventilated patients with COVID-19, but after all the retractions and walk backs, it is unacceptable to tout study results by press release without releasing the paper.”

The U.K. trial is evaluating a number of potential therapies. On June 8, the dexamethasone arm of the study was stopped because its benefit was clear, according to the news release.

It compared 4,321 patients who were randomly assigned to receive usual critical care with 2,104 patients randomly assigned to also receive a low daily dose (6 mg) of dexamethasone in a pill or a shot. The drug, which is widely used to reduce inflammation in arthritis and other disorders, costs less than $1 a day per patient.

For patients on ventilators, the drug cut the risk of death by a third, from 40% to 28%, suggesting one death would be prevented by treating eight ventilated patients. For patients on oxygen, the risk of death was cut by a fifth, from 25% to 20%, meaning one death would be prevented by treating about 25 patients. There was no benefit in patients who didn’t require oxygen.

“It is fantastic that the first treatment demonstrated to reduce mortality is one that is instantly available and affordable worldwide,” study coleader Martin Landray, a physician and epidemiologist at the University of Oxford, said in the release.

“Dexamethasone should now become standard of care in these patients,” declared study coleader Peter Horby, a specialist in emerging infectious diseases at Oxford.

Other experts were more cautious, both because the data have not been subjected to “peer review” and published, and because of the history of corticosteroid use. The drugs were found to do more harm than good in many patients with the flu or previous coronavirus diseases SARS and MERS. Among other things, steroids suppress the immune system, raising the risk of secondary infections and possibly reducing the body’s ability to fight the original infection.

“Steroids have been studied for decades in critical care patients,” said Christopher Seymour, a critical care specialist at the University of Pittsburgh Medical Center. “Up to this point, many of us who have lived through the ups and downs have held off” using it to treat COVID-19.

Seymour called the results “tantalizing” because the British trial is well-designed and the largest to date. “But,” he added, “there is quite a bit that can go wrong by adopting practice changes based on a headline.”

In a tweet, Laurie Garrett, the Pulitzer Prize-winning science journalist, called the results “exciting” but added, “I’ve interviewed many SARS 2003 survivors that have suffered permanent side effects from steroids,” including bone loss and pain.

How might dexamethasone work to relieve severe COVID-19? The news release did not speculate. But experts suggested the drug’s anti-inflammatory and immune-suppressing activities may help the lungs while preventing a potentially deadly immune system overreaction that can be triggered by the worst coronavirus infections.

“You’d like to treat the inflammation to help the lung status, but you don’t want to compromise the body’s ability to clear the virus,” said R. Phillip Dellinger, a critical care specialist at Cooper University Hospital in Camden. “I think what this trial will do is swing the pendulum toward using dexamethasone, pending peer review of this study. Several journals have been stung even after peer review. And we haven’t even gotten through peer review on this.”

Indeed, two of the world’s most respected journals this month retracted papers that were based on flawed statistics from a little-known database of electronic medical records. A Lancet paper linked use of the malaria drug hydroxychloroquine to higher mortality in COVID-19 patients but backpedaled after revealing journal reviewers were not given full access to needed data. (The drug is no longer recommended for COVID-19 patients, as there is no proof it prevents or cures infection.)

And the New England Journal of Medicine said it could not vouch for evidence in an article that concluded certain hypertension drugs are safe for those COVID-19 patients.

Many scientific manuscripts on the coronavirus have come out as “preprints,” published online prior to peer review. And some companies developing vaccines have been faulted for touting early results in news releases, sending their stock values up, only to see them tumble when more complete information came out.