Why aren’t therapeutic antibodies being used more to treat COVID-19?

By Nicholson Price, Rachel Sachs, Jacob S. Sherkow, and Lisa Larrimore Ouellette

When former President Donald Trump contracted COVID-19 in fall 2020, he was treated with monoclonal antibodies, touted as potentially miraculous treatments. Unlike other treatments so touted, there is some rigorous evidence to support these assertions: antibody drugs look like the best treatments currently available to prevent COVID cases from progressing to hospitalization. But months later, the drugs are in limited use and seem to be only a moderately important part of the COVID-19 response. Why aren’t antibodies making more of a difference for ordinary Americans?

What are therapeutic antibodies, and how are they being developed for COVID-19 treatment?

Antibodies are a type of protein produced by immune cells in response to infection. Unlike broad-spectrum drugs, antibodies’ utility lies in their specificity: they target only a minute spot on a foreign invader. This specificity makes antibodies workhorses of molecular biology research. Scientists can also engineer antibodies—to make them pharmaceutically tolerable, for example—for use as therapy. This approach has been wildly successful. In 2019, six of the top ten selling therapeutics in the U.S. are antibodies or fragments of antibodies.

Researchers have been developing numerous antibodies that may treat COVID-19 by mimicking the body’s immune response to thwart infection. To date, there are at least 21 anti-COVID-19 antibody treatments either in or about to begin clinical studies. Two have already received Emergency Use Authorizations (EUAs) from the FDA: Eli Lilly’s bamlanivimab and Regeneron’s cocktail of casirivimab and imdevimab. These two antibody treatments have received substantial interest from the federal government, which has pre-ordered 950,000 doses from Lilly and 300,000 doses from Regeneron. Recently, the FDA issued potency guidelines for future antibody development.   

Recent clinical trial data suggests that antibodies may be quite effective in treating early cases of COVID-19. One treatment from Lilly—a combination of the company’s bamlanivimab and another antibody, etesevimab—appears to decrease COVID-19 hospitalizations by 70% in some measures, a crucially important resource given the disease’s crushing effect on the health care system. Regeneron’s antibody combination treatment had similar results. At the same time, the EUA issued for Lilly’s antibody treatment was for bamlanivimab, alone, and predicated on a smaller patient population, complicating a robust assessment of their ultimate real-world impact. Some, consequently, have criticized the data for single antibody treatments as being “soft” in comparison to that for antibody cocktails.

In addition, a recent, nagging question is how well the authorized antibodies—or those currently in development—will treat new variants of the virus that causes COVID-19; the very specificity of antibodies may limit their effectiveness against changed targets. Eli Lilly’s antibody seems to fare poorly against the South African variant (B.1351), though testing is underway. Preliminary evidence suggests that Regeneron’s cocktail, at least, retains efficacy against the South African and UK (B.1.1.7) variants.

What is limiting the administration of antibody treatments?

As of January 19, states had received over 550,000 doses of Lilly’s and Regeneron’s antibody treatments, but only about 30% of available doses have been administered to patients. Given the promising results described above and the record-breaking case loads over the past month, why haven’t more doses been used?

The problem isn’t affordability. Because of generous federal coverage, including for uninsured patients, price does not appear to be a significant barrier to the use of antibody treatments in the United States so far—which is important given the socioeconomic disparities in COVID-19 infections and deaths. Nor is the problem insufficient manufacturing—at least not yet. States have doses available and more on the way, and the federal government has contracted for 2.5 million doses so far. Rather, as with the COVID-19 vaccines discussed in our most recent post, the immediate hurdle is administration to patients. But the administration problems for antibody treatments are very different from those for vaccines.

Administering an antibody drug to a patient is complex and time-consuming. As explained in instructions to providers for both Lilly’s bamlanivimab and Regeneron’s casirivimab and imdevimab, the drugs need to be infused via IV for at least 60 minutes and then the patients must be monitored for adverse reactions, which requires trained professionals. And the drugs are not authorized for patients who are already hospitalized or who require oxygen therapy—rather, they will only be administered if patients at high risk of severe COVID-19 come to the infusion site for the lengthy infusion process within 10 days of symptom onset, before hospitalization becomes necessary. Those infectious patients also must be sequestered from non-COVID patients requiring IV-administered drugs, many of whom are immunocompromised. 

Navigating these distribution challenges is daunting. Hospitals are swamped with patients who have already developed severe COVID-19, leading many hospitals to refuse their antibody allocations. Testing delays and lack of coordination between testing sites and hospitals mean that patients who might benefit are often not identified in time. States also have differing financial resources for implementing distribution programs, raising important equity concerns. Scarce resources are more likely to be funnelled to triaging the sickest patients rather than treating patients earlier in the disease course. 

These failures in distribution mean that patients often need to advocate for themselves to receive access to the antibody treatments. But many patients don’t know that they need to, or how to do so. Few patients think they should go to the hospital to get an IV drug to keep them out of the hospital. Even resourceful patients who have sought antibody treatments have often had difficulty accessing them, and these access concerns loom even larger for disadvantaged and marginalized communities.

How can policymakers promote access to distributed antibody treatments?

As with vaccines, developing a novel health technology isn’t the end of the process: distributing and administering the new technology to patients in need is critical. Policymakers must consider whether investment in the infrastructure for distributing and administering the product is also necessary, in addition to any investments they have made in the development of the technology itself. Policymakers ought to keep in mind two particular issues as they consider antibody administration: reimbursement and transparency.

Policymakers should think carefully about different strategies for using reimbursement to encourage antibody administration, not merely distribution. Early on, CMS quickly recognized the need to ensure that Medicare would pay many different types of providers to administer these antibodies, and that patients should not pay anything out of pocket for this experience (even if the latter principle was legally more questionable to impose at the time). But given the challenges that resource-constrained state and local officials are facing, additional grants to ensure that facilities have the incentives and infrastructure to administer these products may be an important part of the strategy as well. 

Transparency and coordination will also be critical to improving patient access to distributed antibodies. Many newly diagnosed COVID-19 patients do not know that these treatments are available, and even patients who do know about them and ask their providers about them often have not been able to access them. Patients must be provided with information about how to obtain access to these new therapies, but this type of information could be provided in many different ways. Perhaps people who receive a positive test could also receive information about accessing antibody treatment along with that positive test. Or perhaps states could publicize information about how to obtain these treatments, along with their public information about vaccine distribution and access. 

Policymakers at the state and federal level might also consider encouraging private sector actors to develop novel distribution pathways to lower the logistical barriers to accessing care, some of which is already occurring. As part of their clinical trial, Eli Lilly retrofitted RVs as mobile infusion clinics to allow them to reach residents of nursing homes more easily. Several chains of dialysis clinics, which have the ability to provide outpatient infusions, have begun to administer the new drugs. Pharmaceutical firms may also be encouraged to invest in developing versions of these drugs which are easier to administer, such as inhaled antibody products. 

The rapid development of therapeutic antibodies, like the rapid development of effective vaccines, is a triumph of biomedical innovation. But ultimately, even the best products won’t help unless they make it to patients.  Ensuring the last step is just as critical as facilitating the steps that come before.

This post is part of a series on COVID-19 innovation law and policy. Author order is rotated with each post.

Irene Calboli: Trademarking COVID

Irene Calboli, professor of law at Texas A&M University School of Law, has written a new empirical article reporting that "the COVID-19 pandemic has led to a veritable tsunami of trademark applications" for COVID, QUARANTINE, SOCIAL DISTANCING, and other pandemic-related terms.  Indeed, Calboli concludes that trademarking of COVID-related terms is unusually rampant, even when compared to past tragedies and disasters such as 9/11, Hurricane Katrina, and Ebola. "Ultimately," Calboli writes, "probably only the HIV/AIDS pandemic can be compared with COVID-19 in terms public awareness, societal fears, and strong emotions," and even "the beginning of the HIV/AIDS pandemic did not see a similar rush to trademark filings" that we have seen with COVID. While several practitioners have also noted the large number of applications filed for these terms in the past month, Calboli's paper is the first to survey these applications comprehensively and in detail. 

The full article, Trademarks and the COVID-19 Pandemic: An Empirical Analysis of Trademark Applications Including the Terms “COVID,” “Coronavirus,” “Quarantine,” “Social Distancing”, “Six Feet Apart,” and “Shelter in Place, can be downloaded on SSRN.  The data are included as an appendix to the paper.  The article will be published in the Akron Law Review's IP Symposium Issue as part of the Annual Akron Law IP Scholars Conference. (The other scholars publishing in this year's IP issue are Margaret Chon, Robert Merges, Kristen Osenga, and Sharon Sandeen.) 

I interviewed Professor Calboli on her findings. The interview is transcribed in this post.  To follow along, here is one of the many beautiful tables and informational graphics in the article, included with Calboli's permission.

Table 1  

As you can see, Calboli reviewed a total of 782 trademark applications filed through the end of the year 2020 and presently under examination at the U.S. Patent & Trademark Office. Importantly, a large majority (542 applications) are filed under the Section 1(b) "bona fide intent to use" filing basis, rather than based on actual current use. These applicants will eventually need to convert their applications by proving they are actually using the marks in commerce. So they are in a sense betting on what they will be able to sell and on what will be popular in future.

CAH: Wow, this an amazing paper. It's so comprehensive. I am fascinated by what it reveals about trademarks, as well as about society. The most common term by far was the term COVID itself. You found 519 filings for COVID. The distant second was QUARANTINE, with 121 filings. But CORONAVIRUS meanwhile represented only 54 applications.  Why might this be? Why not more for coronavirus itself?

IRENE CALBOLI: This could be because COVID has become the more common referent in public conversation, not "Coronavirus." It's also possible that applicants thought of "coronavirus" as too generic.

CAH: Around 209 filings were for actual medical products, like face masks, hand sanitizers, and items related to vaccines. Some of them make what look to me like health and safety claims, such as COVID SHIELD for air filters, COVID-19 PROOF for antibacterial cleaning products, and CORONAVIRUS FREE for medical compression tights... Can you speculate on the health and safety implications of this?  Should a trademark be allowed for a mark like COVID-19 PROOF for cleaning products?

IRENE CALBOLI:  I don't think the use of these terms for medical products is necessarily problematic. For example, a mark like COVID-ONESWAB registered for swab tests seems fine to me, so long as it's truly for a test and is effectively able to identify the source of the products. The problem is when you start to see filings for marks like COVID-19 PROOF.  Nothing can be one hundred percent "COVID PROOF"! The Section 2(a) deceptive matter bar should "kick out" and prevent registration of some of the more misleading marks. These applications should be barred as deceptive, since they can mislead consumers. Applicants seem to try to use these signs to market products as "safe" when this simply cannot be proven at all. Hopefully, the examiners will be tough on those sorts of marks. It will be interesting to see what happens as more office actions are written on these applications. It will be especially interesting to review these office actions. 

CAH: A great number of applications were for unrelated products. For example, you found applications for COVID SAUCE for coffee or QUARANTINE& CREAM for body cream. And a ton of filings (around 200) are actually for merchandise and promotional products, T-shirts, hat, pins, and the like. At the Akron Scholars event, you said,

“The million dollar question is why you would want to call your beer COVID, when we all want to move on. But I don’t think people think through that. And we already have a beer called Corona."  

Can you speculate on that? Do consumers really want to buy a HERO COVID-19 2020 T-shirt or a SIX FEET APART IS SMART lapel pin?  If not, why are people still filing these? 

IRENE CALBOLI: I think this dataset is (unfortunately) the perfect example of a case study on the phenomenon of "registering sensation." People swarm in to register a buzz word like COVID or QUARANTINE or SIX FEET APART.  They somehow think they can make money by registering these names and selling products carrying these names. But they don't really seem to think through this. And they don't seem to understand what the proper functions of trademarks are, or what type of signs can, and should, actually be protected. They just want to register new and hot buzz words. Some applicants write in the description of goods/service in the applications language such as "I hope I will sell products with this mark" or "I hope others will not be able to use it." The reality is that people seem to be running to register trademarks that are very likely to be rejected, because they may be descriptive, generic, or misleading, and in turn are unlikely to make the applicant any profits. In fact, these applications are a cost for the applicants who have to pay the fees.

CAH: This is an empirical piece, so you don't necessarily speculate on whether these filings might be a problem as a policy matter.  But are they? Aside from the safety issue we just discussed, is there anything wrong with people getting tons of trademarks on COVID or pandemic-related slogans for T-shirts and bracelets?   Is this a competition problem or a free speech problem? Will these marks "clog" or "clutter up" the system, detracting from the universe of useful marks, as discussed by Barton Beebe and Jeanne Fromer in Are We Running Out of Trademarks?

IRENE CALBOLI: Not necessarily, in my view, because  lot of these marks will be denied and, even if granted, will be very difficult to enforce in practice. Like Rob Merges said at the Akron panel, we witness an interesting phenomenon. We have 519 iterations of the word COVID in the 519 applications filed for this term. Clearly some of these will be redundant and denied, for example under Section 2(d) likely to be confused with a prior mark. However, the large number of filings also maybe casts doubt on the assumption of "language depletion" and its relationship with trademark applications (more studies on this should be conducted).

 

In the paper I do point out that in many instances, terms like COVID and QUARANTINE are used in ways that are quite arbitrary with respect to the associated products, so they may pass distinctiveness under Abercrombie. (Examples in the paper of "arbitrary" uses include COVID SURVIVOR filed for playing cards, and COVID BLUE filed for the advertising of real estate!)

However, there are other serious problems with these applications. In particular, many of these terms and phrases "fail to function" as a trademark. They are only ornamental or informational, they are not functioning as source-identifiers.  As I explain in the paper, the Trademark Manual of Examining Procedure (TMEP), used by examiners, says “failure-to-function” occurs when a mark does not “identify and distinguish the source of a good or service.”  Frankly, a lot of these slogans could fit within this definition. They seem (to quote the TMEP) to "merely convey an informational message" and so are not registrable. So if that is true, I am not sure these COVID trademarks are going to generate serious barriers to commercial activity or speech. 

[The “failure to function” issue is discussed at length in Alexandra Roberts' article in Iowa Law Review called Trademark Failure to Function.  Laura Heyman has also written a wonderful 2019 JOTWELL post on Roberts’ article, called, What We've Got Here is a Failure to Indicate.]

CAH: I agree a lot of these applications seem like they'll be rejected on the grounds you identify. I hope so: There could be a big problem for health and safety if misleading marks like COVID PROOF are granted. But even if they are denied, are there broader costs for the trademark system here?

IRENE CALBOLI: Yes, I think so. The reality is that sifting through all these applications can take up a lot of time and resources on the part of the USPTO.  The costs of examining these marks will be high and do not seem to be justified by any benefits.  Each application is assigned a different examiner on a random basis. It will take many months in order to examine all the applications. This is costly. And at the end, a large number of applications will probably be rejected. After all, the rate of rejection is almost one third for trademarks generally and it will certainly be higher for this group of applications. Also, even if an applicant gets a registration, it is unlikely that s/he will be able to protect the mark broadly or to turn it into a truly valuable brand like. In short, a lot of these signs are not really protectable, and the cons of putting these applications through the process, in my view, outweigh the pros for everyone. They cost the applicants. They cost the system.

CAH: How could the USPTO improve how it deals with these sorts of sensational filings?

IRENE CALBOLI: I would like to see support, even subsidy programs for SMEs (small and medium-sides enterprises) and individuals like in other countries, for good and useful trademarks, but not for the "registering sensation" phenomenon  we are seeing with COVID.  I think that more needs to be done to limit bad filings--filings that are likely to be rejected, do not function as marks, and costs both the applicants and system.  I think a lot of this has to do with education of the public about what can actually be trademarked and what is the function of a mark. 

The system is designed to be user friendly. This is a positive aspect of the system. But the USPTO could do more to deter these filings. Being user friendly does not mean that we need to be so permissive in letting in "junk applications." Now, one of the recurrent problems is that there is a lot of pressure for all IP offices to show high number of filings. And the "junk applications" still count as filings! However, if we spend more resources to examine them than the benefits we get from them (in terms of functioning marks) what is the point? Rather than counting how many applications are filed, we should focus primarily if not exclusively in counting how many marks are actually registered, and how many of these marks are renewed and used in the long term. In other words, despite the pressure of showing "a lot" of applications or how fast applications are processed, the USPTO should focus on how many applications are registered. 

The system can, and should be user friendly, but should do more to sift our bad filings and ideally prevent them. This is not a simple task. For example, some applicants submit fake specimens, because the office says "submit a specimen of how you're using the mark..."  But again, the more the USPTO educates the public about what trademarks are for and what can be trademarked, the more these sensation-driven filing could perhaps be deterred.

CAH: What are your future plans for this project after the Akron Law Review article comes out?

IRENE CALBOLI:  I will continue tracking these COVID-related applications to see their fate at the office. I plan to review the whole dataset in a few months and examine the office actions and the rate of rejections and registrations. I'll be asking questions like: Are these applications denied or granted? if they were rejected, what were the legal basis for these rejections? It will be interesting to see how many of these applications will survive and even if the marks are registered, how many will still be in use beyond the first couple of years?

Also, I would like to expand on this in a new project. In particular, I want to look at Class 25 filings for clothing, apparel, and the like.  I want to look at the rate of rejection in Class 25 and then compare this to rejections in other classes, like say pharmaceuticals (Class 5).  To do this, I need to carefully look at the USPTO data set as well as build ad hoc datasets like I did this time. Another project I am finishing is a similar empirical study on filings that include the words "Black Lives Matter." We have reviewed over 50 applications so far, many filed since Spring 2020, and I am creating a detailed data set of these applications.

CAH: Thank you for all your time, and for such a wonderful read. I really enjoyed this and look forward to seeing it in print!

What can policymakers learn from the disastrously slow COVID-19 vaccine rollout?

By Lisa Larrimore Ouellette, Nicholson Price, Rachel Sachs, and Jacob S. Sherkow

In the middle of a record number of COVID-19 infections and deaths—and continued evidence of racial disparities in the pandemic’s effects—December brought some good news to the fight against the pandemic: the FDA’s emergency use authorizations (EUAs) for the Pfizer-BioNTech and Moderna vaccines and the ensuing nationwide rollouts. The record-breaking vaccine development timeline and the videos of healthcare workers receiving early vaccines are worth celebrating. But the subsequent distribution has been tragically slow—echoing distribution challenges for COVID-related goods ranging from PPE to diagnostics. On Dec. 29, Dr. Leana Wen noted that at the initial vaccination rate, it would take 10 years to vaccinate the roughly 80% of Americans needed to achieve herd immunity. According to Bloomberg’s vaccine tracker, as of today, fewer than 10 million vaccines have been administered in the United States. States are still sitting on nearly two-thirds of the doses they have received, and the federal government is holding half the U.S. supply in reserve. In this post we explain what went wrong and how policymakers can correct course for COVID-19 and avoid such disasters in the future.

Why has the U.S. vaccine rollout been so slow?

We have previously discussed challenges to scaling vaccine manufacturing, which is an important factor in the rollout of these vaccines. But right now, the rate-limiting step is not manufacturing—it’s distribution and administration, including maintaining the cold chain, training healthcare workers on how to thaw and prepare doses, setting up vaccination clinics, and administering the distributed doses to eligible people. States (following the recommendations of the Advisory Committee on Immunization Practices) have typically chosen to offer the first doses to health care personnel and residents of long-term care facilities. In theory, this is supposed to be the simplest stage of the vaccine rollout—as Dr. Saad Omer noted, “[t]his is the part where we’re supposed to know where people are.” These populations are identifiable and reachable in ways that the general public are not—and so vaccine distribution and administration will only get more difficult from here. 

Although there are many factors driving our slowness in distributing and administering the vaccine, two problems loom large: a lack of resources and a lack of federal coordination. States and local governments were allocated just $340 million to prepare for the vaccine rollout, an amount that is highly inadequate. As the Commonwealth Foundation noted, funding is needed for states to “set up cold storage systems, public education and partnership initiatives, arrangements for local vaccine administration with diverse populations and conditions, transport to move the vaccine to the points of administration required to reach all Americans, and information systems to track who has received particular vaccines, and critically, to make sure recipients return for required second doses.” Hiring and training additional staff, purchasing the specialized freezers needed for cold chain management—these important tasks all require funding. And with public health departments already stretched thin by nearly a year of combatting the pandemic—particularly when coupled with the current surge in cases—states needed more resources earlier on.

But an even more significant problem has been the lack of coordination from the federal government, which Dr. Ashish Jha has called a “failure of national leadership.” To be sure, the CDC did send vaccine distribution guidance to public health officials in all states and the largest U.S. cities at the end of August, telling them to prepare for vaccine distribution beginning as soon as late October or early November. But more was needed.

Consider a problem that was entirely foreseeable (and foreseen): getting consent to administer the vaccine from residents in long-term care facilities. Because many residents may not be able to make their own medical decisions (due to the presence of dementia or Alzheimer’s disease, for instance), facilities must often track down the resident’s designated medical decisionmaker, often a family member or attorney, and ask that decisionmaker for consent. Repeating this process for a large fraction of a facility’s residents is an onerous one that can take days or weeks. As a result of this time lag, advocates argue that this process should have been started even before the vaccines received FDA authorization for marketing. But the federal government did not engage in the interagency coordination that would have been needed to do this ahead of time. As a result, even though HHS Secretary Azar publicly said that all nursing home residents could be vaccinated by Christmas, in many states vaccination in long-term care facilities did not even begin until after Christmas.

What can policymakers do to improve COVID-19 vaccine distribution?

To speed up vaccinations—perhaps the only solution to ending the pandemic—there are several targets for policy interventions. First, money. After incessant delays, Congress has finally passed, and President Trump signed, a new stimulus package that includes coronavirus relief. That package includes $8 billion for vaccine distribution (not including $20 billion allocated for vaccine purchases to make them “available at no charge for anyone who needs it”)—a marked increase from the measly $340 million allocated earlier. The federal government could do even more, such as increasing Medicare reimbursements for those paying for the vaccine. For anyone balking at these expenditures, it’s worth pointing out that the pandemic is adding $10 billion per day to the deficit and that around 2.4 million Americans contracted COVID-19 between authorization of the Pfizer-BioNTech vaccine and the relief package’s enactment—roughly 40,000 of whom will die. 

The second policy intervention is improving coordination. To begin with, the federal government—which, for better or worse, has controlled vaccine allotments—should provide more than a few days’ notice of vaccine shipments. Dr. Ashish Jha has suggested the federal government should “provid[e] more direct logistical support” for vaccination sites, including “ensuring that they had the physical infrastructure, the staffing and the IT infrastructure they need to proceed.” This includes providing more guidance to local authorities regarding when they should proceed to the next priority phase, what to do with soon-perishing but unused doses, and—as has been debated recently—expanding distribution within a particular phase. The state of Illinois, for example, recently announced plans to lower the age cutoff for its Phase 1B distribution from 75 years old to 65—in contravention of federal guidelines. Part of the motivating factor behind the drop was equity: because Black and Hispanic Illinoisans tend to die of COVID at younger ages than their white counterparts, Governor Pritzker made the move to “overcome the structural inequalities that have allowed COVID-19 to rage through our most vulnerable communities.” The move also comes against a backdrop of the Governor’s sharp criticism of federal vaccine-coordination efforts.

Third, and relatedly, for those individuals eager to get a vaccine who are not in one of the initial priority groups, state and county health officials could provide substantially more transparency regarding vaccine distribution, including where vaccinations are likely to be administered; how the state or county plans on moving from one priority phase to the next; and how potential patients can go about staying informed about when they will be eligible to receive the vaccine. The sum total of public information from the Los Angeles County Department of Public Health, for example, says only “Some groups in Los Angeles County are already being vaccinated….It is likely to be available to the general public in Spring/Summer 2021. Please talk to your doctor or sign up for email updates to find out when vaccine will be available to you.” Even without guaranteed allotments of vaccines or good coordination from the federal government, content-less statements like these will do little to augment the public’s confidence in getting a shot—or figuring out how to do so—even when available.

Fourth, policymakers should immediately start working on addressing vaccine hesitancy. Right now, it appears that demand outstrips supply, but at some hospitals 40% or more of frontline healthcare workers have declined a shot. Vaccine hesitancy will also make achieving herd immunity difficult: a December survey by the Kaiser Family Foundation found that 27% of the public probably or definitely would not get a COVID-19 vaccine even if free and deemed safe by scientists. Among Black adults, this number rises to 35%, including a fair portion who say they don’t trust vaccines in general and that they aren’t confident that COVID-19 vaccine development has taken the needs of Black people into account. Overcoming the history of medical exploitation of Black Americans should be a particular policy priority, especially given the dire difference in COVID-19 mortality between Black and non-Black Americans and high-profile incidents such as the recent death of Dr. Susan Moore.

How can the federal government improve vaccine distribution institutions going forward?

The first step to improvement is to recognize that vaccines are not vaccinations, and that each step of developing and deploying a new biomedical technology is costly. Operation Warp Speed provides a successful model of how the public sector can speed development and manufacturing by, among other things, de-risking costs, facilitating collaboration and regulatory approval, and providing knowledge infrastructure. But if the government is going to take full charge of the vaccine, it must facilitate dissemination as well. 

While the federal government isn’t the only potential actor here—public health has long been a state bailiwick—the long-term dismantling of state public health infrastructures and the rather disparate state responses to the pandemic, in general, the federal government seems to be the single entity that could (if functioning well) make the biggest difference next time. On that score, the federal government could improve vaccine distribution going forward in at least two different ways: taking direct action on its own or creating incentives for the private sector. 

On the direct front, the federal government has substantial resources that could be devoted to solving “last mile” problems. The United States Public Health Service (PHS; the part of HHS that includes the FDA, CDC, and NIH) also includes the Commissioned Corps, a uniformed service of federal public health employees. Outside the PHS, the military has expertise and experience in logistical operations including large-scale distribution, especially in coordination with the national guard.

An alternative, indirect approach is to focus on a funding structure that better incentivizes vaccination by the private sector. The private sector, including pharmacies, employers, and hospitals, already oversees the administration of many vaccines each year (notably the annual flu vaccine). But in a pandemic situation with limited supply, urgent timing, and complex logistics, careful coordination with manufacturers and others administering the vaccine is essential. Manufacturers could potentially play a coordinating role, but lack incentives to do so. Currently, incentives for development (e.g., for drug manufacturers) are wholly distinct from incentives for administration (e.g., pharmacy chains or hospitals getting reimbursed). The federal government could link these incentives; for instance, it could provide manufacturers a reward per patient vaccinated rather than simply for receiving vaccine approval or manufacturing a certain number of doses.

The optimal approach may be a mixture of both types of incentives, such as a combination of private sector coordination and efforts at a centralized federal infrastructure. Both the public and private sector are subject to failures, and not all eggs should be put in one basket. Coordinated redundancy is not only acceptable in a pandemic, it is desirable. Accordingly, policymakers shouldn’t be afraid of spending too much on vaccine distribution or development—even under low-end estimates of social value, the billions spent on COVID-19 vaccine development so far are only a small fraction of this value. Part of the reason this pandemic has been as devastating as it has been was insufficient vaccine development funding prior to the pandemic. Providing adequate rewards at each stage of the vaccine pipeline will help us do a better job with the next outbreak.

This post is part of a series on COVID-19 innovation law and policy. Author order is rotated with each post.