Those treated with the drug while in hospital intensive care units needed mechanical ventilators to help them breathe up to three days less, on average, than patients who did not receive it, the data showed.
Dexamethasone, which is administered via intravenous injection, first received approval for use in people with endocrine and rheumatic disorders from the U.S. Food and Drug Administration in 2003.
It is one of several corticosteroids -- hydrocortisone is another -- that have shown promise in improving symptoms in people hospitalized with severe COVID-19.
"It is relatively rare in medicine that you find drugs where the evidence of their effectiveness in saving lives is so consistent," Dr. Derek Angus, who was part of a study on the use of hydrocortisone in seriously ill patients, also published Wednesday, said in a statement.
"People on ventilators or oxygen and under intensive care should definitely be given corticosteroids," said Angus, professor and chair of the Department of Critical Care Medicine at the University of Pittsburgh.
The findings build on those of the "Recovery" trial, a British study that found the steroid dexamethasone reduced deaths by about one-third in COVID-19 patients on ventilators and by about one-fifth among people who needed oxygen but were not on ventilators.
A separate study, published in July by the Journal of Hospital Medicine, found that hospitalized COVID-19 patients with high levels of inflammation in their lungs saw a 75% reduction in their risk for death after receiving steroid treatment. However, the study did not specify which steroids were used.
For the new dexamethasone study, researchers at 41 hospitals in Brazil administered the drug to 151 patients with the virus who were being treated in ICUs. Over a 28-day period, those given the steroid had 6.6 ventilator-free days, while patients who did not receive it had 4.4, the data showed.
Dexamethasone treatment did not significantly reduce patient days spent in the ICU, the researchers said. During the study, 85 patients who received the drug died, compared to 91 of the 148 patients who did not, they said.
Meanwhile, the new hydrocortisone study found a 93% probability that giving patients a seven-day IV course of the drug would result in better outcomes than not giving the steroid.
Based on the results of studies to date, the World Health Organization issued new guidance Wednesday recommending the use of corticosteroids for "patients with severe and critical COVID-19" only. The drugs are "readily available globally at a low cost," the WHO said.
"We've found clear, reliable evidence in high-quality clinical trials of how we can tackle this devastating disease," Dr. Anthony Gordon, another co-author on the hydrocortisone study, said in a statement.
"We now have more than one choice of steroid treatment for those who need it most," said Gordon, professor of anesthesia and critical care at Imperial College London.
"Having a choice of different types of steroids, all of which seem to improve patient recovery, is great as it helps ease the problem of drug supply issues," he said.