An old drug can learn new tricks during the coronavirus pandemic.

That’s the main takeaway from the World Health Organization (WHO) in a new analysis of corticosteroids—a class of drugs which have existed for dozens of years and are far cheaper than new, experimental COVID treatments in development—suggesting that drugs like dexamethasone can slash the chances of COVID-19 related deaths by as much as 35% in the sickest patients.

The WHO analysis of coronavirus drugs encompassed seven separate studies. And while an analysis of this sort—what’s called a “meta-analysis”—isn’t as rigorous as other types of trials like a randomized controlled study, the data are compelling.

Corticosteroids have a very different action mechanism from many of the other coronavirus drugs in development. COVID-19 is a peculiar disease. Some who have been infected may be asymptomatic. For others, an active infection may spell misery for months. And for others yet, the misery may linger through their own immune systems’ response in attacking the pathogen.

The new WHO analysis suggests that steroids such as hydrocortisone and dexamethasone, which fight inflammation, can sharply cut the risk of death for COVID patients sick enough to require a ventilator or other extraordinary medical measures rather than just cutting down on the number of days one needs to spend in a hospital.

That doesn’t mean everyone infected with COVID should get such treatments. In new guidelines, the WHO issued a “strong recommendation” for corticosteroid therapy, either via IV or orally, for seven to 10 days for sickest patients. For those without severe illness—not so much at this point.

The recommendations may carry cost considerations, too. Dexamethasone, a generic steroid, can cost as little as six dollars. New treatments for COVID in the middle of an outbreak may well hike the prices of those therapies’ innovation costs for pharmaceutical companies.