Cheap, widely available steroid drugs reduced the number of deaths in the sickest patients with COVID-19, show a trio of newly published clinical trials.

The World Health Organization, citing evidence from these and similar trials, announced Wednesday it strongly recommends doctors use the medications to combat severe or critical forms of disease caused by coronavirus infections.

Finding a treatment that saves lives is “electrifying … it gives us hope. Maybe we’re gaining on this virus,” said Todd W. Rice, a critical care physician at Vanderbilt University Medical Center who was not involved in the studies.

WHO’s decision brings the international agency in line with the U.S. National Institutes of Health, which earlier this summer released guidelines for clinicians to use a synthetic steroid, dexamethasone, to treat hospitalized patients who require ventilators or oxygen.

The evidence that persuaded the WHO included a meta-analysis, sponsored by the organization, which evaluated three new studies, plus four other randomized, controlled trials. Each trial involved a medication from the family of anti-inflammatory drugs called corticosteroids.

“These three trials, and then the World Health Organization meta-analysis, sets steroids as the standard and the expectation that patients are critically ill will get treated with this,” Rice said.

The WHO review concluded corticosteroids reduced deaths in critically ill patients by 20 percent – a drop from 2 deaths per 5 patients to 1 per 3. It found that another steroid, hydrocortisone, had benefits on par with those of dexamethasone.

“Corticosteroids are the only treatment that has been conclusively demonstrated to reduce mortality in patients with COVID-19,” said Jonathan Sterne, an author of the meta-analysis and an expert in medical statistics at Great Britain’s University of Bristol.

“These results strengthen the evidence that doctors should treat critically ill COVID-19 patients with corticosteroids, unless there is a strong reason not to do so based on the circumstances of the individual patients,” he said.

Dexamethasone was the first medication shown to increase the odds of survival in patients with severe COVID-19 when the initial outcomes from a British clinical trial, named Recovery, were published in June.

That these new studies showed similar benefits was a welcome confidence boost to doctors who had been using corticosteroids to treat patients based on Recovery’s results.

“The Recovery trial was a fantastic trial. But even though it was a great trial, it was viewed by many as simply one trial,” said University of Pittsburgh critical care doctor Derek C. Angus, an author of the eight-country hydrocortisone trial. “People wondered about generalizability.”

To steroids’ true believers within the medical profession, it was compelling, he said, but not enough to convince all doctors.

These new trials should help resolve remaining uncertainty. “When we see that this is this very consistent signal of benefit across different trials, that’s very exciting,” said Hallie Prescott, a pulmonary and critical care doctor in the University of Michigan Health System.

The three trials published Wednesday, each in the Journal of the American Medical Association, involved COVID-19 patients in intensive care units from across the planet. One trial studied dexamethasone in 299 patients in Brazil; a second in France tested hydrocortisone in low doses in 76 patients against 73 who received a placebo; the third enrolled 379 patients to test hydrocortisone in the U.S. and seven other countries.

All three trials halted after Recovery’s results became public in June. Trial steering committees around the world concluded it would be uncomfortable, if not unethical, for doctors to give patients placebos in randomized trials.

Because of that, these steroid trials stopped short of enrolling enough patients for their results to be as statistically powerful as desired.

“Although the individual trials weren’t necessarily conclusive, because they hadn’t run to term, they were all pointing in the right direction,” Angus said. He added: “It feels like the first fresh piece of fairly consistent, uncomplicated, straightforward, good news about something concrete to do for the sickest of patients.”

Similar results in dexamethasone and hydrocortisone indicate corticosteroids are helpful as a class of drugs. Such flexibility may allow hospitals to avoid the shortages of remdesivir, an antiviral coronavirus therapy. The NIH recommends treatment with the steroid methylprednisolone, for instance, if dexamethasone is not available.

“The steroids are pretty much readily available and they’re not really expensive,” Rice said. That’s true not only in the U.S. but in lower-income countries, where the WHO guidelines will carry significant weight.

While providing a strong foundation for corticosteroid treatment, the studies left several questions unanswered. One of the biggest is more clearly determining which patients are most likely to be helped.

“This sets forth that the sickest, the most severely ill in the ICU appear to be clearly benefited by this,” Rice said. Even within that group, however, there may be certain patients for whom “giving steroids isn’t the right thing to do.”

Doctors are likely to agree an ICU patient on mechanical ventilation or significant respiratory support should receive steroids, Prescott said. But she said she will make case-by-case decisions for her “gray zone” patients – those who receive a few liters of oxygen.

“Just because steroids work in sicker patients does not mean to say that we should be cavalierly starting steroids in every patient,” Angus said. Many patients “don’t need these big-gun drugs that suppress the immune system broadly.”

Steroids may not help low- or moderate-risk patients, and some doctors worried about potential harm in those cases. The Recovery trial did not observe a benefit in patients with mild covid-19.

“We always have a concern when we give steroids to patients in general, because steroids depress the immune system,” Nahid Bhadelia, medical director of the special pathogens unit at Boston Medical Center, told The Washington Post in June. “They carry the risk of making the current infection worse and increase the chance of another infection taking a foothold.”

Steroids do not attack the virus itself. Instead, they decrease the body’s defensive response, which in severe coronavirus cases can spiral into out-of-control inflammation. Whether steroids should be given in combination with other treatments, such as remdesivir, that directly fight the pathogen is another open question.

“We don’t know that for sure, because we don’t have rigorous studies showing us that,” Rice said.

Steroid drugs can also cause unwanted side effects, such as delirium or high blood sugar levels.

Future studies will refine duration and drug dosages, the doctors said. A trial in Denmark has already been launched, Prescott said, to evaluate high versus low doses of steroids.

In the halted trials, some coronavirus patients received tapered treatments, meaning doctors reduced the amount of steroids at time went on. In other cases, patients were given fixed dosages for seven days.