Remdesivir was the first treatment given Food and Drug Administration (FDA) emergency authorization for treating critically ill patients with COVID-19. Several trials—including a randomized study conducted in collaboration with the National Institutes of Health (NIH)—have shown that remdesivir can reduce the amount of time very sick patients have to spend in the hospital. But whether or not it reduces the risk of dying from COVID-19 is way more complicated.

In its latest release, Gilead says that its study shows that remdesivir may lower mortality risk for coronavirus patients by 62%. That figure, however, is an assessment of relative risk, which is different from absolute risk.

Without wading too deep into the mathematical weeds, what this generally means is that comparing percentages of patients who die with one therapy versus another treatment course isn’t the be-all and end-all of a treatment’s effectiveness. You can lower a death rate from 10% to 8%. That’s technically a 20% relative reduction; but on an absolute level, you still have 8% of people dying.

In Gilead’s case the company claims that for patients being treated with remdesivir, after 14 days the mortality rate fell to 7.6% versus 12.5% for those being treated with existing standard-of-care options.

That’s a noticeable difference. But the nuances persist, as stated by experts in Gilead’s own press release. The study was not randomized (meaning, it pitted different data pools together, which can lead to discrepancies), and there wasn’t a placebo arm to test true efficacy on the mortality front.

That doesn’t mean that it’s completely insignificant information. But it does mean that a whole lot more data in more rigorous clinical settings, along the lines of the NIH trial showing remdesivir can at least cut down the recovery time for severely ill COVID patients, would be required.

“While not as vigorous as a randomized controlled trial, this analysis importantly draws from a real-world setting and serves as an important adjunct to clinical trial data,” said Dr. Susan Olender, of Columbia University Irving Medical Center, in Gilead’s statement.

Former FDA commissioner Scott Gottlieb echoed the sentiment about balancing hope and scientific caution in a tweet. “This is very encouraging but needs to be confirmed in a prospective trial,” he wrote, referring to the need to conduct a trial that isn’t based on analyses from disparate sources.

More coronavirus coverage from Fortune:

• Why Black-owned businesses were hit the hardest by the pandemic
• Pop-up retail was made for the pandemic
• How the coronavirus crisis has affected female founders
• The enduring history of health care inequality for Black Americans
• E-book reading is booming during the coronavirus pandemic