Cross-posted from Written Description, where it originally appeared on April 7, 2020.
The escalating pandemic has caused devastating shortages not only of ventilators and personal protective equipment like masks, but also of essential medicines needed to treat COVID-19 patients. As detailed by STAT and the New York Times, prescriptions for painkillers, sedatives, anesthetics, and antibiotics are up, but the rate at which prescriptions are filled and shipped to hospitals is down. The FDA helpfully tracks drug shortages, but this doesn’t solve the problem. With the sudden spike in hospitalized patients with COVID-19 symptoms, physicians are using these drugs faster than manufacturers are making them.
What is causing these drug shortages?
Drug shortages are frighteningly common even in the best of times. A 2019 FDA report noted that from 2013 to 2017, at least 163 drugs went into shortage. (The actual number is likely much higher.) That report blamed “economic forces”—namely, price-eroding generic competition, a lack of incentives to make quality generic manufacturing more efficient, and supply chain difficulties that made the continued manufacture of older generics unprofitable. These problems are now exacerbated by the sudden demand spikes caused by COVID-19 patients. As just one example: propofol, an important drug for sedating patients who need intubation—and, historically, already in waxing and waning states of shortage—has seen prescriptions shoot up about 100%.
Supply has been slow to meet COVID-19-related demand—but slower still because of the outbreak’s disruption to the global supply chain. Many pharmaceutical ingredients are manufactured in China, which has seen slowdowns (and in some cases, shutdowns) in manufacturing sectors across the country. Furthermore, because drugs do expire, they’re not stockpiled when there’s a surplus. In some instances, countries have banned the export of drug products important for treating COVID-19 to ensure adequate supply for their own citizens. India, for example, has banned exports on hydroxychloroquine in the event the drug proves useful in treating COVID-19. It’s a wicked problem: the very thing causing the sudden spike in demand is shutting down the means of supply.
Importantly, it should be noted what the problem isn’t: patents. There have been recent calls to “break” pharmaceutical patents, both in the U.S. and abroad, in view of the public health necessity for some medications and their consequent short supply. But patents have not—either in general or for COVID-19—caused the shortages of important drugs. Rather, the issue arises in the face of generic entry, where generic competition is so intense that over time it has made the manufacture of a drug unprofitable. Unlike, say, Daraprim and Martin Shkreli, drug shortages aren’t a problem of greed; they’re a problem of aligning manufacturing incentives at the right time.
What could the US do to increase drug supply in the short term?
The best short-term strategy for increasing the availability of these drugs is to expand the production capacity of existing facilities. Manufacturers may take these steps independently, but if not, Secretary of Health and Human Services Alex Azar could use the authority of the Defense Production Act (delegated to him by executive order on March 18) to expand capacity as well, at least for domestic firms. The federal government should also remove the legal roadblock created by Drug Enforcement Administration quotas for domestically manufactured controlled substances, as the American Hospital Association and other medical groups have already asked it to do. (DEA quotas focused on decreasing opioid production—including several products now in shortage—may be exacerbating the problem.)
To be clear, even with federal encouragement, expanding production capacity is not an immediate fix. Experts suggest that increasing domestic production may take two to three months, but current COVID-19 models suggest that the supplies will still be needed then.
In addition to expanding the production capacity of existing facilities, the US could credential new facilities or new manufacturers. But this approach would be even more time consuming due to approval and inspection requirements. (Although the FDA has postponed most foreign and some domestic inspections in light of the pandemic, at present they are still conducting pre-approval inspections, both foreign and domestic.)
In lieu of these above approaches, other strategies may be used to expand capacity more rapidly. The FDA has already permitted compounding pharmacies to engage in bulk manufacturing of hydroxychloroquine, and they might extend the same permission to some of these additional products. But this would be a limited solution: many of the drugs at issue are administered by IV or injection, making them more complicated to manufacture and increasing the importance of sterility. In 2012, there was an multistate outbreak of fungal meningitis caused by tainted steroid injections from the New England Compounding Center; policymakers need to balance the benefits of getting needed drugs more quickly against this kind of risk.
Congress also recently passed, and the President enacted, the CARES Act, which among many things, provided FDA with additional authority to respond to drug shortages by expediting approvals and inspections and tracking decreases in drug supply. But unfortunately—and for reasons wholly unclear—the drug-shortage provisions of the CARES Act do not become effective until September 23, 2020, almost certainly too late to have any effect on this first wave of the coronavirus epidemic.
How can the US manage demand for drugs during these shortages?
Because even the best short-term strategies for increasing drug supply take time, over the coming weeks the US will have to make do with a limited supply. Once COVID-19 patients are in critical condition, they often need certain medications to help stay alive; for example, hospitals are experiencing shortages of the sedatives necessary to put patients on ventilators. But the US can attempt to manage demand by slowing the growth in new patients and preventing stockpiling.
“Flattening the curve” illustrations have widely circulated over the past month to show how protective measures can delay and reduce the outbreak peak to stay within health care system capacity. And that capacity includes not just hospital beds and ventilators, but also drugs needed for hospitalized patients. Maintaining shelter-at-home orders remains critical during the slow ramp up in drug manufacturing and boosting medical surge capacity.
In addition to slowing the growth in new COVID-19 patients, the US can promote policies to prevent stockpiling of drugs that do not fill an immediate need. As illustrated by US grocery shelves, where panic-buying has created more shortages than disruptions to grocery supply chains, stockpiling for future use can create scarcity even if there is enough supply to satisfy immediate demand. Pharmacy benefit managers are taking steps to limit pharmaceutical hoarding, including limiting coronavirus treatments to a 10-day supply. State lawmakers or pharmacy boards could impose similar policies, thus taking into account public considerations rather than purely private ones. As we mentioned last week, they’ve already done so in the context of hydroxychloroquine.
What long-term changes should be made to the global drug supply chain?
The COVID-19 pandemic highlights fundamental flaws in the global drug supply chain; hopefully we can use this as a wake-up call to address these problems before the next pandemic.
Most fundamentally, the global supply chain is old, haphazard, and fragile. Drug companies rely almost exclusively on batch manufacturing processes rather than the continuous manufacturing processes seen in many other industries. This makes it hard to nimbly respond to changes in demand and generally decreases flexibility. And if something is wrong with a batch, the whole batch needs to be tossed. A move to continuous manufacturing would lead to easier ramping up and an easier time constantly monitoring quality. The FDA has pushed for this, and the 21st Century Cures Act includes funds to ease the move, but firms have been slow to change.
Behind this reluctance are several factors that reduce innovation in drug manufacturing. Most fundamentally, changes in manufacturing require FDA review, and incentives to improve manufacturing in the first place are attenuated—among other things, manufacturing process patents are often difficult to enforce. The FDA’s 2019 report notes the effects in grim detail. The market doesn’t help—most people have no idea who manufactures most of the drugs they take, especially generic drugs or drugs administered in hospitals, so the market creates little incentive to compete on quality. (The FDA has, however, suggested a new way to let manufacturers signal high-quality manufacturing processes.)
Increased transparency in the process of drug distribution and use would also help. Right now, drugs are distributed through a complex network of middlemen, and firms often don’t know how fast the drugs are being used until the next order comes in, which may be too late for them to scale up capacity to meet increased demand. Sharing more data, as close to real-time as possible, would help firms know what’s coming.
Finally, the FDA approves processes for producing each drug, which means that it’s hard to tweak and change those processes to improve them and keep them robust over time. The agency could change that to let trusted manufacturers certify changes to their processes, allowing them to innovate more easily. This would still demand oversight, and would be challenging to implement, but the FDA is already piloting such an approach, including change management, in the digital health space, so the idea has precedent.
Right now, drug manufacturing is a relatively fragile ecosystem that involves a lot of old technology; that makes it especially vulnerable to shocks, and less responsive to spikes in demand. Increased innovation and information could help change that landscape so that as the next health crises emerge, drug manufacturers can be ready.