The Covid-19 pandemic has laid bare multiple weaknesses in supply chains for goods important to Americans, including medical products—both those to prevent and treat Covid and those for sundry health conditions. According to the Food and Drug Administration, the number of drug and biologic shortages increased steeply before the pandemic, from a low of 41 at the end of 2017 to 76 at the end of 2019, then grew to 86 by the end of 2020. The total number of drug and biologic shortages reached 125 on February 16, a number approaching all-time weekly highs of 140 in 2012 and 2013.
Though the FDA’s reports of growing shortages seem to accord with our broader experience of supply chain disruptions, another fundamental problem remains: the FDA’s methods to identify and characterize shortages are free from basic economic principles and therefore essentially incoherent. This failure creates a blind spot for policymakers and the public alike. As the Biden administration and Congress look to strengthen pharmaceutical and medical-product supply chains, it’s imperative that the FDA develop a better way of measuring the severity and length of product shortages.
Currently, the FDA identifies shortages—periods when demand exceeds supply at prevailing prices—using opaque and arbitrary criteria. The agency states that it uses information from manufacturers and distributors and market-share data to establish whether a shortage exists, without explaining how it determines the existence of excess demand. The approach does not comply with information-quality guidelines issued in 2002 by the Office of Management and Budget, which stipulate that agencies “disseminating influential scientific, financial, or statistical information” should provide “a high degree of transparency about data and methods to facilitate the reproducibility of such information by qualified third parties.” Because the FDA plays a key role in tackling drug shortages, its conclusions that shortages exist and the information it uses to make such determinations are surely influential.
Unfortunately, FDA determinations are anything but transparent and reproducible. The FDA reported preventing 199 shortages in 2020, the most since 2012, and up from 126 in 2016. But given that the FDA has made it impossible for the public to understand just what—in the agency’s view—constitutes a shortage, its claim to be preventing more shortages is unverifiable.
It does not have to be this way. We know from private reports of shortages that the FDA could easily make useful information about drug shortages available to Congress and the public. In 2020, following a public report that a corticosteroid called dexamethasone was effective in treating severe cases of Covid-19, Vizient, a large group-purchasing organization, reported daily wholesale measures of demand and supply. Daily excess demand jumped to roughly triple the numbers of tablets or vials typically sold. In addition, aggregate demand rose week over week by 610 percent, and the fill rate—the percentage of ordered units that can be filled using existing inventory and anticipated production flows—dropped from 97 percent to 54 percent. (The fill rate allows calculation of excess demand as a percentage of total demand; a fill rate of 54 percent means that excess demand was 46 percent of total demand.) These estimates, available just days after data collection, likely described market-wide conditions, because Vizient reported using data from half of the nation’s acute-care providers and 20 percent of ambulatory-care providers.
The FDA’s reports to Congress of drug shortages do not mention quantitative measures of demand and supply or fill rates. The reports thus obscure crucial differences in severity. And they leave open whether the agency is responding to shortages based on economic data like demand and supply or other metrics.
Examples help to illustrate the arbitrariness of the FDA’s measurement. Take the current shortage of morphine sulfate injection, for which the agency lists four manufacturers and 33 different products. One manufacturer, Hospira Inc, discontinued five products in November 2021 but continued to produce eight others. The duration of this shortage—more than four years—suggests a cause more complicated than the “demand increase for the drug” noted on the FDA’s website. But because the FDA provides no data on excess demand in its determination that the drug is in shortage, we have no sense for the severity of the shortage or whether it is any closer to resolution than it was four years ago.
By contrast, the FDA now treats a shortage of an oral anti-depressive, sertraline hydrochloride, which began in May 2020 and lasted to March 2021, as “resolved.” The market for the drug involves five manufacturers making 73 solid oral products in various package sizes, and two manufacturers making four liquid oral products. The agency’s website does not report on the original cause of the shortage—for example, whether it stemmed from manufacturing quality-control problems or an unexpected increase in demand. It doesn’t shed light on the severity of the shortage or how it was resolved—whether by the agency’s prioritization of regulatory decisions to allow marketing of comparable products or changes in medical practice.
When the demand for a safe and effective medical product goes unmet, it is important to understand why. That FDA reports are virtually useless in trying to understand the causes, severity, and duration of a drug shortage is inexcusable from the standpoint of public health. As the Biden administration looks to strengthen medical supply chains, Congress should insist on better information from the FDA as part of its oversight responsibilities.
The FDA could easily use data on fill rates to develop consistent and objective criteria for identifying the beginning, end, and severity of shortages. It might decide, for example, that shortages occur whenever weekly demand exceeds weekly supply by more than 10 percent of supply, or whenever weekly excess demand exceeds historical averages by more than three standard deviations. While no single measure will be perfect, any systematic effort to provide clear and systematic economic data on shortages would represent a huge improvement over the status quo. What’s more, such reporting will inform the development and evaluation of the policies necessary to ensure the resilience of medical supply chains.