Why are hospitals struggling so much with vaccine distribution?

A huge logistical effort collides with all of hospitals' existing challenges and then some

I wrote an article for the American Prospect, out today, outlining why Operation Warp Speed was so successful on the vaccine development but so slow on the vaccine roll-out. I hope you’ll read and subscribe to the Prospect, their coverage of COVID and the rollout especially has been great.

And now here we are, a month after the first vaccines began shipping for non-trial distribution.

Source: https://www.bloomberg.com/graphics/covid-vaccine-tracker-global-distribution/

The U.S. is doing well in terms of sheer vaccine doses delivered, but it’s falling behind on percentage of population vaccinated. Why are we struggling to keep up?

Staff, stuff, space, and systems

In Paul Farmer’s recent book Fevers, Feuds, and Diamonds, an excellent account of the 2014 Ebola epidemic in West Africa, he talks about the need for staff, stuff, space, and systems to manage any pandemic. Broadly, the U.S. is suffering from weaknesses in all four right now. Staff and space are overwhelmed with COVID-19 patients, leaving little capacity for vaccinations. The stuff—vaccines—is lagging behind demand as manufacturers try to keep pace. And systems at all levels were caught unprepared for this distribution.

Already, several entities responsible for distributing the vaccine are struggling. As I wrote in November—and as I talk about in more depth in the Prospect piece—there was a risk to relying solely on huge players CVS, Walgreens, and McKesson for the vaccine distribution. That risk is showing itself; CVS and Walgreens are struggling to meet vaccination targets in long-term care facilities, where they are responsible for providing the bulk of shots. The two states that have the highest percentage of distributed vaccines in the arms of residents right now are West Virginia and North Dakota, both of which relied primarily on independent pharmacies for the long-term care facility rollout.

Other states are starting to turn to new entities to help with the rollout. When CVS fell behind schedule, Oklahoma called up the National Guard to distribute the vaccine to elderly veterans. Washington State announced that it was making plans with Starbucks and Microsoft, both multinational companies with home bases in WA. It’ll be interesting to see if other states continue to make moves like this, and if Biden, his administration officials, or prospective HHS secretary Xavier Becerra start looking to smaller, more nimble entities to distribute the vaccine.

But for this newsletter, I want to talk about the friction points for hospitals. Why are they struggling so much with vaccine distribution?

Getting to full vaccination

For this rollout, there are four steps for health systems.

First, health systems have to build trust with their patients. This could be as simple as sending a text, communicating with patients that vaccines are available, and that they will be eligible soon.

Next, hospitals have to proactively reach out to patients who are eligible and provide a platform for scheduling the first dose. This is a complicated and heterogeneous process. In some states and localities, like DC, both health systems and local governments are simultaneously handling this outreach and scheduling.

In DC, for example, patients who are in the GW Hospital system have been receiving periodic updates through FollowMyHealth, a patient platform.

Simultaneously, the DC government has been using more of a concert ticket model. Every few days, the government sends a mass email to DC residents, notifying them that, say, 1,400 vaccination slots for residents aged 65+ are opening at 9am the next morning. The slots are first-come, first-served.

Third, health systems have to follow-up with patients to schedule the required second dose of the Pfizer and Moderna vaccines. Some systems are choosing to schedule the appointments upfront, at the time that the patient receives the first dose. But some are relying on patient contact at the 3- or 4-week mark, when the patient is scheduled to come in for the second dose.

Finally, health systems have to administer and monitor the second dose. Pfizer and Moderna’s second doses seem to cause a pretty intense immune reaction—far preferable to taking your chances with COVID, but still, getting the second dose can mean a day of chills and fatigue. Some health systems are monitoring side effects proactively; others are being deluged with calls from worried patients who haven’t been adequately warned about what a normal response constitutes.

All of these logistical challenges are conflicting with the very structure of hospitals. Hospitals are designed for short-term, bounded interactions and emergency care. Rounding doctors have time to see patients for maybe 2 minutes at a time. Nurses are similarly stretched.

This is all compounded by the strain of treating COVID-19 itself—hospitals are struggling under the weight of too many COVID patients, complicating a smooth rollout.

The challenges of an equitable distribution

Anecdotally, it seems that hospitals’ level of commitment to a smooth vaccine rollout is closely correlated with their level of commitment to the community. Larger academic medical centers and chain hospitals, without a close connection to the surrounding community and with the majority of their revenue coming from complex and expensive procedures, aren’t prioritizing investment in the infrastructure and staff needed for the vaccine rollout.

To give one example, UPMC, which has tight control over Pittsburgh but whose leadership has seemed to care more about revenue than dispensing sound medical advice during the pandemic, has prioritized giving vaccines to its employees—even those who work from home—over UPMC long-term care facility residents or elderly Pittsburghers. Other hospitals just aren’t building the systems they need to adequately communicate with patients and provide doses quickly.

But even with the best of intentions, hospitals have struggled with the key issue of equity in distribution.

In DC, for example, many of the early vaccinations went to people in the wealthiest sectors of the city. This isn’t necessarily because of corruption; more likely, it’s that higher-paid health care workers live in those parts of the city, as do people with more access to information.

But it’s still a disparity, even if an unintentional one. So now DC is trying to rectify it by only offering vaccinations to certain lower-income zip codes that have received fewer vaccinations per capita—a worthy effort that further increases the burden on registrars and health care workers as they struggle to keep track of who is eligible for which doses.

A similar attempt to prioritize access in communities of color in Dallas ground to a halt this week after state officials told Dallas leaders that they would cut off vaccine supply if Dallas proceeded with prioritizing certain zip codes.

Why is this so much harder than the flu vaccine?

Hospitals seem like obvious vaccine partners—after all, many hospitals do wide-scale flu vaccination campaigns every year. Why is this harder?

First, hospitals so far simply haven’t known how much vaccine supply they’re getting ahead of time. With the flu vaccine, they have months to plan and order a precise number ahead of time.

Next, the segmentation of who is eligible for the vaccine (health care workers, nursing home residents, all those 65+, smokers, etc.) is making the process endlessly complicated. Not only do hospitals have to figure out what segmentation strategy their locality is following (and the strategy has been varying wildly), they also have to communicate that to patients and find and invite the correct population segments to receive the vaccine. With a flu vaccine, hospitals can advertise for months, and then give it to whoever shows up.

Relatedly: flu vaccine campaigns are now a common feature of the public health advertising landscape. As with anti-smoking engagements, it’s now commonplace to see ads for flu vaccines on bus stops. It’s generally considered good citizenship to get a flu vaccine. And while COVID vaccines should be viewed similarly, there’s still a level of public mistrust that extends beyond flu vaccine skepticism. There hasn’t been a public confidence campaign yet, and so a lot of education and communication is falling on individual providers.

Dr. Kimberly Manning with an example of how to communicate vaccine confidence with patients—it takes time and empathy, a big but essential ask for exhausted health care workers.

Finally, the need for first and second doses is compounding all of these logistical issues. Not only do hospitals have to track and schedule and inject patients for one dose—they have to do it twice! This is a logistical challenge that would be stunning for any entity. Hospitals are over here doing it with fax machines and Windows XP.

What does this mean for the future?

Well, a combination of the, ahem, change in administration and health systems’ growing familiarity with the vaccination process seems to bode well for an accelerating vaccine rollout.

Forward-looking hospitals are investing in technical and physical infrastructure, and the staff needed to run them. As those systems come online and become more operational, forward-looking hospitals are well-positioned to handle the bulk of vaccinations that will take place when eligibility is open to all.

But as I talk about in the Prospect piece linked at the top, a logistical effort of this size and scale requires central, flexible planning teams at the national (and local) levels. Hospitals and hospital trade groups should be pushing for more government involvement and funding—and they should also be building their own internal central planning teams.

As for forward-looking changes, it appears that COVID will do for hospitals’ digital infrastructure what it has done for telehealth more broadly: accelerating need and, simultaneously, adoption. And as hospitals get dragged further into the digital age, it may be that we see growth in their digital literacy and willingness to implement new systems. Rather than going with one vendor for decades, hospital staff may be more willing to seek out and acquire innovation—they may even need to get used to a new tech stack.

Finally, it may very well be that community hospitals come out ahead. By building out their infrastructure and demonstrating to local patients their commitment to health, community hospitals are showing themselves to be superior in care to less-engaged chain hospitals and lofty academic medical centers. The next time a large system tries to acquire the community hospital, patients may think twice before going along.

Thanks to those who gave me thoughts on this topic!