U.S. Hospitals Need to Study How Well Remdesivir Really Works

For months, hospitals have suffered the consequences of the country’s pandemic blunders. Intensive-care units have overflowed. Medical staff have struggled to get personal protective equipment, and have worked beyond exhaustion. It seems unfair to ask them to make up for another government misstep. But U.S. hospitals need to band together to run a clinical trial of the Covid-19 treatment remdesivir, even though the Trump administration has already decided the drug is what all hospitals should be using, going so far as to corner the worldwide market for it.

Under the government’s deal with Gilead Sciences Inc., hospitals pay $3,200 per course of treatment for the antiviral drug — which is about 33% more than governments of other countries are to be charged. (The lower price will be available only to Veterans Affairs, the Indian Health Service and a handful of other U.S. government insurers, covering about 4% of the population.)

The decision to commit so profoundly to remdesivir was based on evidence that was flimsy to begin with and is now outdated. One study showed that remdesivir improved Covid-19 recovery times and might reduce mortality in hospitalized patients. Another smaller study weakly supported the recovery-time benefit, but not the mortality one. (The data that Gilead released on Friday provided no additional insight into remdesivir’s potential benefits.) (Disclosure: I received a lecture fee from Gilead Sciences in 2017.)

Both studies were published in May, a lifetime ago in pandemic years. In the spring, the coronavirus was targeting the elderly: In the first week of March, 60% of people admitted to the hospital with Covid-19 were 65 or older. The average age of subjects in the two studies was 59 and 65, respectively. In the first study, only 11% were under 40. 

By the last week of June, in contrast, 72% of patients admitted to hospitals with Covid-19 were under 65, and 42% were under 50. And antiviral treatments have varying risk-benefit ratios across ages. This is why the Centers for Disease Control recommends antiviral influenza treatment only for patients over 65 or under 2, absent other disease complications.

There is every reason to suppose that the risk-benefit calculus for remdesivir will vary by age, too. Florida data show that hospitalized patients 65 and older have a 44% risk of death — nearly nine times that of people 25 to 44 (whose risk is 5%).

What’s also changed since the remdesivir studies were done is the standard of care for Covid-19 patients. Doctors have learned that the common steroid dexamethasone reduces the risk of death for the severely ill by 20% to 35%. It’s not known whether remdesivir will work as well, or work at all in patients receiving dexamethasone.

U.S. hospitals can answer that question and many others by following the example of the dexamethasone study. It was run across the U.K.’s National Health Service hospitals, with some patients randomly assigned to standard care, some to dexamethasone, and some to other promising therapies.

Doctors throughout the U.K. were encouraged to enroll patients in the study, called RECOVERY, which involved simply collecting basic information such as their age and presenting complications. Random assignment to the various treatment arms ensured balance in all other factors that could affect the results.

Ultimately, 15% of patients in Britain hospitalized with Covid-19 were enrolled. At the time the dexamethasone finding was published, the study already had 11,000 participants — 10 times as many as were included in the one study suggesting remdesivir is beneficial, and about five times as many as would be needed in a study to answer the question that the U.S. government already considers settled: Which patients do better — those receiving dexamethasone, or those receiving dexamethasone and remdesivir?

The U.S. has no such collective approach to studying Covid-19 therapies. For American hospitals, accustomed to competing with one another, cooperating on a large scale would be a sea change. Only a few hospitals routinely coordinate clinical trials.

But note that even the most conservative models project that tens of thousands of people will be hospitalized for Covid-19 this month, and most models put the number in the hundreds of thousands. With patient participation at even a fraction of what of the U.K. hospitals accomplished, a remdesivir study could be fully enrolled within days, and provide an answer within weeks.

The hospitalization numbers are a reminder of a central tenet of clinical research: A study of very few can improve the well-being of very many. Because we are in a crisis, it’s tempting to skip this needed study and just assume remdesivir is more likely to help than hurt. But that could be wrong, and lead to more unanswered questions. What will we do when there is another treatment that might help people? And what about the one after that? Treatments can work synergistically, but they also can counteract one another. The only way to know how to use them is to study them separately and together.

A study like this is also the most ethical way forward. The July supply of remdesivir is projected to be 92,000 treatment courses, far less than what hospitals will need. Many hospitals have run out already. So not all patients will be able to get the treatment, and randomization is a good way to decide who does. If shortages are to be chronic, that’s all the more reason to figure out which people benefit from remdesivir, and which do not.

Since the early days of the pandemic, it’s been obvious that our hospitals need to keep up with scientific findings about new treatments. Now it’s apparent that they also need to run some of the studies themselves.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Peter B. Bach, a physician, directs the Drug Pricing Lab at Memorial Sloan Kettering Cancer Center in New York.

©2020 Bloomberg L.P.

BQ Install

Bloomberg Quint

Add BloombergQuint App to Home screen.