Getty / The Atlantic

Editor’s Note: The Atlantic is making vital coverage of the coronavirus available to all readers. Find the collection here.

Updated at 10:23 a.m. ET on September 29, 2020.

On the day that a COVID-19 vaccine is approved, a vast logistics operation will need to awaken. Millions of doses must travel hundreds of miles from manufacturers to hospitals, doctor’s offices, and pharmacies, which in turn must store, track, and eventually get the vaccines to people all across the country. The Centers for Disease Control and Prevention, along with state and local health departments, coordinates this process. These agencies distributed flu vaccines during the 2009 H1N1 pandemic this way, and they manage childhood vaccines every day. But the COVID-19 vaccine will be a whole new challenge.  

“The COVID situation is significantly different and more complex than anything that we have had to deal with in the past,” says Kris Ehresmann, an infectious-disease director at the Minnesota Department of Health.

The two leading vaccine candidates in the U.S.—one developed by Moderna, the other by a collaboration between Pfizer and the German company BioNTech—have progressed so quickly to clinical trials precisely because they are the fastest to make and manufacture. They rely on a novel vaccine technology whose advantage is speed, but whose downside is extreme physical fragility. These vaccines have to be frozen—in Pfizer/BioNTech’s case, at an ultracold –94 degrees Fahrenheit, colder than most freezers—which will limit how and where they can be shipped. The ways these vaccines are formulated (without added preservatives) and packaged (in vials that hold doses for multiple people) also make them easier to develop and manufacture quickly but harder to administer on the ground.

In other words, speed is coming at the expense of convenience. “For this first generation of vaccines, we won’t trade off safety. We don’t want to trade off effectiveness,” says Kelly Moore, the associate director of immunization education at the Immunization Action Coalition. So instead, the U.S. is planning for a vaccine that requires brutally complicated logistics. Public-health departments in states, territories, and major cities are currently drawing up vaccine plans for the end of October. It’s still unclear whether these vaccines are safe and effective—and it’s extremely unlikely that data will be available by the end of October. But the departments are getting ready. Many are already stretched thin by the ongoing pandemic, and they are now helping plan, as Moore puts it, “the largest, most complex vaccination program ever attempted in history.”


The leading vaccine candidates both deploy a new, long-promised technology. Their core is a piece of mRNA, genetic material that in this case encodes for the spike protein—the bit of the coronavirus that helps it enter human cells. The vaccine induces cells to take up the mRNA and make the spike protein and, hopefully, stimulates an immune response.

By using mRNA, vaccine makers do not need to produce viral proteins or grow viruses, methods that are used in more traditional vaccines and that add time to the manufacturing process. This is why Moderna and Pfizer/BioNTech have been able to get their vaccines into clinical trials so quickly. Moderna went from a genetic sequence of the coronavirus to the first shot in an arm in a record 63 days.

To get a naked strand of mRNA inside a cell, scientists have learned to encase it in a package called a lipid nanoparticle. mRNA itself is an inherently unstable molecule, but it’s the lipid nanoparticles that are most sensitive to heat. If you get the vaccine cold enough, “there’s a temperature at which lipids and the lipid structure stop moving, essentially. And you have to be below that for it to be stable,” says Drew Weissman, who studies mRNA vaccines at the University of Pennsylvania and whose lab works with BioNTech. Keep the vaccine at too high a temperature for too long, and these lipid nanoparticles simply degrade. Moderna’s and Pfizer/BioNTech’s vaccines have to be shipped frozen at –4 degrees and –94 degrees Fahrenheit, respectively. Once thawed, Moderna’s vaccine can then last for 14 days at normal fridge temperatures; Pfizer’s, for five days.

The freezer temperature required by Moderna’s vaccine makes it difficult to ship; the ultracold temperature required by Pfizer and BioNTech’s vaccine is nearly impossible to maintain outside of a large hospital or academic center with specialized freezers. For this reason, Pfizer has devised “thermal shippers” that, unopened, can keep the vaccines frozen for up to 10 days; once opened for the first time, they have to be replenished with dry ice within 24 hours, then every five days. These shippers are supposed to be opened no more than twice a day to take out vials, and must be closed within one minute. The real catch, though, is that these shippers hold, at a minimum, 975 doses of the COVID-19 vaccine.

A large hospital in a city could deal with that volume, but in rural areas, a 975-dose shipment will need to be broken up into smaller ones—all while making sure the vials stay ultracold. “The other potential would be only shipping that vaccine to our more urban areas,” says Molly Howell, North Dakota’s immunization program manager, “but then we’re leaving out a lot of people who are health-care workers in rural areas or at high risk in rural areas.” To get the vaccine out to those places, her department is looking into buying frozen-transport coolers and potentially a dry-ice machine. If North Dakota is allocated, for example, 2,000 doses, the state will have to open the thermal shipper, repackage smaller allotments in dry ice, and physically drive them to rural clinics across the state. The vaccines are too precious to risk shipping conventionally.

The storage and handling requirements for these vaccines are especially stringent, but they’re also especially uncertain. In time, it may turn out that these mRNA vaccines can be stored at higher temperatures or can be reformulated to be stored at higher temperatures, as other vaccines have been. Scientists are actively trying to create more stable lipid nanoparticles, and Pfizer says it is working on a freeze-dried version of its vaccine that can be kept in normal freezers. These incremental improvements in storage are a normal part of the vaccine-development process, but they take time. For example, Kathleen Neuzil, a vaccine researcher at the University of Maryland School of Medicine, points out that the flu vaccine FluMist initially needed to be frozen but can now be stored at normal fridge temperatures. (Neuzil is also an investigator on the Pfizer/BioNTech-vaccine trial.) In the August CDC meeting where Pfizer unveiled the thermal shipper, a CDC official interjected to tell stakeholders not to go out and buy freezers in anticipation of a vaccine. The agency was exploring other storage solutions, and the requirements could change.

And in fact, between that August meeting and the publication of the CDC’s vaccine-distribution playbook in September, the number of days Pfizer and BioNTech’s vaccine could be stored at fridge temperature increased from one to five. The vaccine is so new that even its manufacturer is still figuring out its minimum storage requirements.


It sounds absurdly simple, but how the mRNA vaccines are packaged also imposes logistical challenges. Currently, they’re in multidose vials that have to be used or discarded within six hours of opening. Moderna’s vaccine comes in 10-dose vials; Pfizer and BioNTech’s, in five-dose vials. Unused doses can degrade over time at high temperatures and, more dangerous, can become contaminated with bacteria, because the vaccines lack preservatives. Both the multidose vials and the lack of preservatives help get a vaccine out faster, says Moore: Experts have been worrying about a shortage of glass for vaccine vials, and preservatives add complexity that can slow down vaccine development.

Vaccine providers in the U.S. are unaccustomed to giving multidose, unpreserved vaccines, though. Administering them will require scheduling appointments with extra care in order to minimize waste, but also discarding unused doses if needed for safety. When multidose vaccines are used outside the U.S., according to Moore, who chairs a World Health Organization immunization committee, some waste is built into the vaccination program. “It’s okay to open a vial for one baby,” she says, because a program that doesn’t waste any doses is probably erring on the side of turning people away. But this mindset might seem counterintuitive, especially while COVID-19 vaccines remain scarce.

Lastly, both Moderna’s and Pfizer/BioNTech’s vaccines require two doses per person over time, and the second dose has to come from the same manufacturer as the first dose. It also has to be administered 28 days, for Moderna’s, or 21 days, for Pfizer and BioNTech’s, after the first dose—in both cases longer than the vaccines can be stored in the fridge. All of this means that having the right number of vaccines for the right people will require extensive and careful record keeping.

Individual states maintain electronic immunization registries that track which residents have gotten which vaccines. What needs to be reported to the registries varies state by state; many vaccine providers, such as pharmacies and pediatrician’s offices, directly connect their records to the registry. But doctors who don’t routinely give vaccinations, such as those who see adults, might not be connected, which could mean manually inputting the data for every patient into the immunization registry. New connections to the system can also take weeks or months to establish, because of the complexity of electronic health records, Moore says.

The CDC is rolling out a new Vaccine Administration Management System (VAMS) to supplement existing state registries, and it is expected to offer features such as scheduling and supply management. But VAMS has also added confusion, says Rebecca Coyle, the executive director of the American Immunization Registry Association. One issue is that the CDC is now requiring identifiable information that some states are not allowed to share from their existing immunization registries. If that’s not reconciled, vaccine providers might have to spend hours manually inputting that patient data into the new VAMS. “There’s a lot of clarification that still needs to happen,” Coyle says. “The clock has started with states to finalize their response plans, and yet there are giant chunks of information that are missing.”

The two-dose requirement for these vaccines also runs up against the problem of human nature: People forget. They can’t get off work. They can’t find child care. They might even move. “That’s just normal human behavior outside of COVID,” says Azra Behlim, a senior director at the health-care-services firm Vizient. The CDC is planning to send physical vaccination-record cards for each patient along with vaccine supplies when states order their doses. It is also encouraging public-health departments and hospitals to send reminders about a second dose. This is important because even a full course of a vaccine may offer only partial protection against COVID-19, and one dose is likely to offer even less.

If a vaccine is fast-tracked through an emergency use authorization rather than formally licensed by the Food and Drug Administration, that too could create bureaucratic hurdles. For example, Medicare doesn’t cover the costs of emergency-use drugs. So while the government intends to pay the cost of the vaccine and of supplies like syringes, hospitals would be on the hook for storage, scheduling, record keeping, and paying staff to actually give the injections. “Hospitals are not happy about that—at all,” Behlim says. A fix will likely have to come from Congress.


Another worry for hospitals: having to juggle multiple vaccines that are not interchangeable, especially after more become available in the future. “What they’re concerned about is: I get a vaccine now in November, and then another manufacturer launches in January, and then another manufacturer in March, and three more launch in May,” Behlim says. Immunization registries can record who got which vaccine, but hospitals and clinics will still have to decide which ones to stock and how much of each. One vaccine might be more effective, but another one easier to store. A third might be most effective in older people, while a fourth could have the advantage of requiring only a single dose. The more vaccines there are on the market, the harder vaccine management becomes.

In fact, with dozens of vaccines currently in clinical trials, the U.S. will very likely have multiple COVID-19 vaccines from multiple manufacturers next year. Two other vaccines are just behind Moderna’s and Pfizer/BioNTech’s mRNA vaccines, in Phase III clinical trials in the U.S. One of those is made by AstraZeneca and the other by Johnson & Johnson; both insert the genetic code for the coronavirus spike protein into a harmless virus.

These vaccines take slightly longer to manufacture, because they require growing viruses, and they are also a relatively new technology. But they do not have to be frozen, and Johnson & Johnson’s can be given in just a single dose. Close behind these two are more traditional vaccines that use proteins purified from the virus, which will likely have traditional storage requirements. Of course, clinical trials still need to be completed before scientists will know whether any of these vaccines are safe and effective. “Which vaccine or vaccines will prove the safest and the most effective and the most deployable? I think we don’t know yet. And that’s why having redundancy is good,” says Dan Barouch, a vaccine researcher at Harvard. (His lab is a collaborator on Johnson & Johnson’s vaccine.)

In the short run, speed is of the essence. But in the long run, these other characteristics—safety, effectiveness, and ease of use—will determine which vaccines get widely distributed. Julie Swann, who studies supply chains at North Carolina State University and who worked with the CDC during the 2009 flu pandemic, says she’s disappointed that the U.S. has put its weight behind these mRNA vaccines, which rely on new technology and whose handling imposes extra requirements on states and vaccine providers. It will be even harder to use them in developing countries. “There’s no way we can use this in some countries around the world,” she says.

The good news is that more deployable vaccines are moving fast through the pipeline too. The race to a vaccine has dominated hopes for an end to the pandemic. But the first COVID-19 vaccine may not ultimately be the most important COVID-19 vaccine.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.