Peter Chin-Hong, MD

When I looked at the WHO guidelines regarding therapeutics and remdesivir use in patients with COVID-19, I was disappointed. I was disappointed not in WHO bringing together a group of experts from around the world who have varying viewpoints and varying resource settings but in the uniformity of the voice in not recommending that remdesivir be used, or that no evidence supported its use in any COVID-19 hospitalized patients.

The evidence that was cited was not new — we all knew about the Solidarity trial, the NIH trial, the Gilead study and the study from China. But they're all very different, and to use that together to generate consensus opinion is almost like comparing apples to oranges. Specifically, the notion that not having a mortality benefit was a justification for not supporting the use of a therapy is problematic for several reasons.

First, mortality widely varies from locale to locale. When you look at the Solidarity trial and you look at all of the countries that were involved, it becomes problematic when you lump them all together in a trial for the use of remdesivir. The lower the mortality is, the harder it is to find a difference between drug and placebo. Speaking from my role in California where our COVID-19 mortality rate in general has been low, if we look for a mortality difference in any of our therapeutic studies regarding COVID-19, we would not find it — just because we do not have enough patients who died overall.

The second point is that the intervention may be too late or not in the right sequence. When you put a hodgepodge of patients together for remdesivir, you're looking at patients in different arches of the disease. We know that remdesivir is an antiviral and probably works best in the early stages of disease, when viral replication is still afoot. Dexamethasone, the anti-inflammatory, probably works best in later disease, when you have more of an exuberant inflammatory response, and maybe the sweet spot is using both and not considering them individually.

The third point about using mortality as an endpoint in COVID-19 is that there are other things that affect mortality in COVID-19 — not just therapeutics. We have talked a lot about surge capacity. You can be in the best hospital in New York during the tristate surge, but your mortality rate is still going to be 20% to 30% in your hospital, compared with a hospital that is equally skilled that hasn't had a surge, where your mortality rate is going to be less than 5% to 10%.

When you use a therapy and you look at mortality, it doesn't really capture all the other aspects that affect mortality, many of which can't be measured easily. That all leads to the use of more proximal endpoints, which is something that people have brought up before in other fields like critical care medicine, where they're not necessarily looking at mortality anymore, but they are looking at your number of days on the ventilator or days requiring oxygen. Maybe it could be a similar framework in which we approach not just whether remdesivir is a useful drug but uniformity across all COVID-19 therapeutic trials.

At the end of the day, it ultimately is disappointing to give a thumbs down to remdesivir because it's going to set up a divide between the haves and have-nots in the world if there's not going to be a global body that advocates for cheaper and more widespread distribution of a drug that could have a benefit, certainly in a population of COVID-19-infected patients. Rather than say, “Let's wait for evidence,” we should use the evidence that we do have. In the best, most rigorously controlled studies conducted by the NIH and others, if you look at other endpoints like number of days and hospitalization, and specific subgroups that may benefit disproportionately compared with others, there is likely a role for remdesivir. To divide the world into the haves and the have-nots is not in anyone's best interest at this moment.