Why Faulty Modeling Is Causing US Hospitals To Run Out of Remdesivir

US government reliance on historical data rather than predictive models and failure to monitor demand relative to supply is causing hospitals in hot zones like Texas, Florida and California to run out of the only drug that so far has improved patient outcomes. JL

Eric Boodman reports in Stat:

Hot spot states are feeling shortages. Clinicians are left to tackle the highest caseloads in the country without one of the few tools proven to work against the disease. Covid-19 inpatients at Houston Methodist Hospital jumped from 200 in May to 700 in July, and the flow of the drug hasn’t kept up. When the hospital got its latest parcel, there was already a waiting list. Using last week’s numbers to determine this week’s delivery, rather than a predictive model, creates a lag in places where the case count was burgeoning. (And) officials are also failing to track the use of the drug, leaving them unable to assess which hospitals have an oversupply and which are running out.

With Covid-19 patients filling her Houston hospital, pharmacist Katherine Perez wishes she had enough remdesivir to treat all those who might benefit.

In June, as Texas businesses welcomed customers and coronavirus cases surged, Houston Methodist, the hospital where she works, restricted the medication to only those patients who weren’t yet on ventilators, in the hope of making the supply last longer.

Even with those narrowed criteria — and with enough vials to treat 200 more patients delivered last weekend — there still isn’t enough. Covid-19 inpatients jumped from around 200 in May to over 700 in July, and the flow of the drug hasn’t kept up. “We are on the verge of running out, maybe today, maybe tomorrow,” the infectious disease pharmacist said on Thursday afternoon.

It would be the second time in two weeks. When the hospital got its latest parcel, the biggest one yet, there was already a waiting list. “We got a call from the National Guard on the Fourth of July, which was already pretty remarkable, to let us know that they were going to drop off the shipment that afternoon,” she said. “It was a nice surprise. … We’ve just been getting patients started from the backlog that we had when we ran out the Wednesday before.”

Related: 

Data show panic and disorganization dominate the study of Covid-19 drugs

Other hot spot states are feeling such shortages, too. That means clinicians are temporarily left to tackle some of the highest caseloads in the country without one of the few tools proven to work against the disease. It’s the latest in a long string of issues plaguing the U.S. government’s distribution of this drug — and one that physicians predicted back in May.

Soon after a clinical trial revealed that remdesivir could shorten coronavirus patients’ recovery, federal officials started sending cases of it directly to medical centers, but with no rhyme nor reason as far as hospitals could tell. Some overwhelmed centers were skipped over, while less burdened places received precious shipments — and physicians decried the unfairness and lack of transparency.

The government tried to course-correct by collecting Covid-19 patient numbers from hospitals, using those data to determine how much of the drug should be sent to each state and territory, and then letting each one’s health department further parcel out its allocation.

That made the calculus clearer — but doctors still worried that using last week’s numbers to determine this week’s delivery, rather than a predictive model, might create a lag in places where the case count was burgeoning.

An empty shelf at Houston Methodist that used to hold the powdered form of remdesivir.COURTESY KATHERINE PEREZ

Now, after distributing all 940,000 vials that Gilead Sciences, the drug maker, had donated to the government, the Trump administration has secured a new supply — enough for about 500,000 patients — that federal officials will divvy up by state and that hospitals will then be billed for. It’s supposed to last through September, but already, with each new shipment, many doctors find themselves playing catch-up.

In Florida, for instance, on July 1, a pharmacy director for the Jackson Health System in Miami wrote an email to physicians saying, “the Department of Health has temporarily exhausted all the remdesivir supply in stock for the treatment of COVID-19. We are scheduled to receive another shipment tomorrow.”

But within four days, the hospital’s shelves were nearly empty again, and the state once again had no more to ship them. “We will complete treatment for patients that have started therapy,” the director wrote on July 5 — implying that those who were just arriving at the hospital were, at least for now, out of luck.

Helen Boucher, the chief of infectious diseases at Tufts Medical Center in Boston and a leader of the Infectious Diseases Society of America, has been hearing similar stories from some of her counterparts in a few states, including Texas, Florida, and Arizona. “As a physician, my job is to offer the best therapy I can to the patient who needs it. Remdesivir is a therapy that demonstrates efficacy,” she said. “It’s one tool in the toolbox, and I would certainly hope my colleagues in Houston have that tool.”

Remdesivir is one of two medications that guidelines suggest doctors should use outside of research studies — and the only one that stops the coronavirus itself, rather than blocking the body’s harmful immune reaction to it. In a government-run randomized clinical trial, the drug was shown to shorten the time for hospitalized Covid-19 patients to recover from 15 to 11 days on average.

On Friday, Gilead released a new analysis from another study of remdesivir called SIMPLE, which did not contain a control group and instead compared the data to a historical control. It said that patients on remdesivir recovered faster and were 62% less likely to die.

But the data are in no way definitive. Scientists tend not to trust such comparisons because they are frequently wrong. And the Gilead analysis shows a much bigger mortality benefit than the main government study that compared remdesivir to placebo, in which mortality was 30% less in the placebo arm than among those who received the drug — a result was not statistically significant, meaning that researchers are not certain it is not due to chance. Gilead said in a news release that the new mortality finding “requires confirmation in prospective clinical trials.”

Even so, having gaps in the remdesivir supply is a serious problem.

“If your loved one was in a hospital in Florida and was dying of Covid and couldn’t get access to this drug, while a patient with mild symptoms in Illinois is getting it, that’s not fair and it’s not the best for the care of patients,” said Michael Ison, an infectious disease physician at the University of Northwestern Medicine in Chicago. “That’s especially so when we’re talking about a federal resource that should be equally distributed across the country.”

To him, the issue is that the federal government is not increasing allocations of remdesivir enough to keep up with the rise of cases in hot spots. He worried that officials are also failing to track the use of the drug in various regions, leaving them unable to assess which hospitals have an oversupply and which are running out.

A Department of Health and Human Services spokesperson said Friday that tens of thousands of remdesivir vials were being shipped out that same day to help with coronavirus hotspots in Texas, Florida, Arizona, and California, and that more would be going to some of them early next week. But she confirmed that one week’s allocation was based on a past week’s statistics. “If state A has 10% of the national total of admitted patients with COVID-19, state A gets 10% of the remdesivir for a given distribution week,” she wrote in an email to STAT, acknowledging that the percentage might have changed by the time the shipment goes out.

The agency, she added, “remains committed to equitable and efficient distribution of the drug with the goal of reaching as many patients as possible across all states and U.S. territories.”

Not every hospital was scraping the bottom of the barrel. “We are monitoring supply and trying to ensure optimal use; however, at this point, we are not having to deny treatment due to lack of drug,” explained David Nix, a professor at the University of Arizona College of Pharmacy, in Tucson. “The supply is not flowing in but we are expecting new shipments on a regular basis.”

For Perez, the pharmacist at Houston Methodist, though, the rationing that her team has to do is “emotionally exhausting.”

The criteria her hospital established for receiving such a precious drug are meant to direct it to those patients early enough in their sickness to benefit the most from an antiviral. But she also knows that it might still do some good for those who were just intubated, too — if only her team had enough for that. “Every bone in your body wants to give them remdesivir,” she said.