Rushing a COVID-19 Vaccine Could Lead to 'Perpetual Limbo'

; Abraham Verghese, MD; Natalie E. Dean, PhD


September 14, 2020

Find the latest COVID-19 news and guidance in Medscape's Coronavirus Resource Center.

This transcript has been edited for clarity.

Eric J. Topol, MD: Hello. I'm Eric Topol, and with me is my co-host, Abraham Verghese, for a new edition of Medicine and the Machine. We have with us Natalie Dean from the University of Florida. She has a background as a biostatistician and vaccinologist, and she worked with the World Health Organization on Ebola and Zika responses. Now, to me, she's a guiding light in the COVID-19 pandemic. Welcome, Natalie.

Natalie E. Dean, PhD: Thank you so much for having me.

Topol: Can you give us your overall sense of where we've been with COVID-19 since January, where we are today, and where we're headed?

Dean: We're struggling right now in the United States. Certainly, we had a terrible spring; we were able to turn things around to some degree, but then we slipped and our numbers are back up. We are at a weird state, with over 1000 deaths on average per day. Yet people's reactions to it have really changed. We're heading into this fall, with schools and colleges reopening and people back at workplaces. Yet we don't know what's going to happen with the seasonality of the virus. There's just a lot of uncertainty about what's going to happen over the next few months.

Topol: Uncertainty certainly is a theme. You've always been upbeat, asking what we can do to make things better. Do you ever lose that sense along the way? Do you ever get a sense of defeatism?

Dean: It's difficult to see how things have played out, knowing that there's so much more that we can do. Early on, I listened to a podcast with Jim Kim from Partners in Health. He talked about the early days of the AIDS epidemic, when it was exploding in Africa. Problems can seem insurmountable, but you just need to get started and do what you can. So even when the problems feel overwhelming, whatever little bit I can do to help, that's what I want to do.

Topol: Let's hear your plan for how we get out of the pandemic in the United States. What do you think it's going to take?

Dean: It's going to take a coordinated response that uses all of the features that we know work. They may not be exciting or new, but things like contact tracing and more widespread testing. Certainly, I'm excited about any innovations in rapid testing. And more mask wearing, more education, and clearer messaging. All of these boring public health things that people don't always appreciate, they can add up.

Some places in the country are figuring it out. Massachusetts and New York are figuring out contact tracing. These are systems that didn't really exist before. I mean, they had contact tracing for certain diseases, but not at the scale that we're talking about. To get these things to work, it takes time, and trial and error. But once we have that in place, we can learn a lot about what are high-risk and low-risk settings, and we can have better eyes on the situation and figure out how to keep people safer.

Topol: Can you explain what forward and backward contact tracing means?

Dean: When we talk about contact tracing, traditionally we mean that you identify a case, and then you track down all of their contacts in order to refer them to quarantine or maybe get them tested. They stay home, which prevents future generations of transmission. That's forward tracing. But what's challenging with this virus is that people can infect others before they even have symptoms. So you have to be really fast to catch that next generation.

The idea of backward contact tracing is that you're focusing on trying to find the source of a person's infection. We know that this virus spreads through superspreading events or clusters, so if we look backward, it's possible that by finding the source we will find other people infected from the same source.

If you can find those people and get them tested, then maybe you can identify more chains of transmission and catch up with those and track down their contacts. So it's a way of looking backward in order to look forward and find more chains of transmission. It's also a way to understand the types of sources that appear over and over — maybe certain types of settings, like workplace settings, or parties, or events — to figure out what the risky settings are.

Abraham Verghese, MD: I'm intrigued by your background working with Zika and Ebola. How did that inform you as this new virus came along, clearly a very different virus? It seems to me that, even with Zika and Ebola, you weren't dealing purely with a virus and an immune response. You were also dealing with all of these social imponderables in the countries that were involved. Talk about that challenge.

Dean: Ebola is a much more severe pathogen, so we were able to detect more of the cases. There are relatively few asymptomatic or very mild infections. Zika was at the other end of the spectrum, with some severe events with pregnant women and Guillain-Barré syndrome, but then there was also a big pool of asymptomatic infections. With COVID, we're somewhere in the middle. But this is a very large epidemic, affecting so many people.

My research has focused on vaccines. Primarily I've been thinking about how we evaluate new vaccines during pandemics and how we make that work. That has to do with the surveillance systems in place and our understanding of the epidemiology of the virus.

Verghese: You've been quite eloquent about herd immunity, or rather the fallacy of herd immunity. Explain to our listeners what that means and the realities of herd immunity for this virus.

Dean: I think there is a lot of confusion about herd immunity and what that means exactly. We're used to talking about herd immunity in the context of vaccines. In that case, it is a safe way to protect people who are unable to receive vaccine — either they're too young or they're immunocompromised. If enough people receive a vaccine, then new outbreaks cannot take hold, and that protects the people who are not vaccinated.

But in the context of COVID, we're talking about infection-induced herd immunity. When people get infected, they develop immunity, and the idea is that maybe that could protect other people from getting infected.

That's not really a protection strategy because what happens along the way is that all of these people get infected and get sick. So we may see situations where hard-hit places have accrued some level of immunity that will impact future dynamics. But it is still unclear whether we will see another outbreak in New York City, for example.

But what New York City had to go through in order to get to some level of immunity in the population is not something we would want for other places. That's why, when we talk about what we should do next, we're focusing on the proactive strategies we can put in place to protect people.

Verghese: People often point to Sweden as an example. But as you point out in some of your Twitter feeds, that's hardly an example of something we should emulate, at least not for the reasons they are stating.

Dean: Sweden is a complicated situation. It has not always been portrayed accurately in the media. They started with a lighter touch but they have since added protective measures and now what they have in place is, in many ways, similar to what other places have. So they started with a lighter touch and added things, whereas other places started with more stringent measures and backed off.

I have questions about some of Sweden's choices; for example, they're not using a lot of mask wearing, but if you look right now at their numbers, they're not doing poorly. It would be interesting to understand exactly what's going on in Sweden. I think a big part of it is that they have a very compliant population that is adopting voluntary measures that have been recommended by the government. A lot of people also live alone and spend a lot of time outdoors during the summertime. So the lessons learned there may not apply to other countries, particularly the United States, where we don't have the same healthcare system and people may not comply with recommendations.

Verghese: It is an important misconception that Sweden had this "let it rip" story, when in reality they've taken many measures and their numbers are looking good these days.

You talked about the nonpharmacologic interventions — masks and distancing, etc. But the buzz right now is about vaccines, and the potential of prematurely stopping or even overriding the phase 3 trial so that we don't have phase 3 data. What are your thoughts about where we stand with these two US trials that are going to be completing enrollment perhaps by the end of September, but not a lot of follow-up at that point?

Dean: Over the past few years, I've been involved with the World Health Organization's (WHO) Research and Development Blueprint Initiative. Since Ebola, we've been thinking about how to evaluate vaccines during emergencies, and how to make these trials successful during an outbreak in these complex settings.

Over and over, we've come back to this: You need to make sure you get the evidence in order to inform the decisions. Once you put something in place, such as through an emergency use authorization, it's likely that you may not be able to generate any more evidence from that point on. So if you put an emergency use authorization in place prematurely, then you can be left in this state of perpetual limbo where you are not quite sure if the product works.

For example, the WHO has stated, and the US FDA has agreed, that a vaccine should be at least 50% effective. But also we should be able to rule out that the lower bound should not be below 30%. We want to rule out that a vaccine is 20% effective, 15% effective, which would not justify widespread use. One concern I have is that these early looks at the data may not provide enough information to rule out that lower bound. We may see a point estimate above 50%, but are we confident that it's not a low efficacy vaccine? Are we adhering to the guidelines that have been prespecified for the level of evidence we need? If we don't follow those, we may never be able to go back and figure it out.

Verghese: You're bringing out the point that, for example, the Pfizer CEO has talked about having 100 events and having a 50% reduction. But the confidence intervals over those point estimates are all over the place. To extrapolate that to give a vaccine to tens of millions of people seems like a pretty big reach.

Dean: And there are different levels of evidence needed for different situations. If you're talking about using a vaccine in a limited way for postexposure prophylaxis, like a rabies vaccine or anthrax, there are different levels of evidence that are needed. But if you're going to be vaccinating millions of people, the level of evidence you need is really high. It is important to reiterate that we want to be very confident that our product is safe and effective, exactly because we plan to use it in so many people.

Verghese: This is different from the other vaccine programs you've worked with, because now we're talking about, potentially, billions of people getting a vaccine. Also, any given trial will have a limited number of people who are older than 65 years, and they won't include children. What are your thoughts about these different groups for whom we will not have sufficient data?

Dean: That's something that epidemiologist Marc Lipsitch and I have been talking about lately. We know that these trials are designed to look at the overall trial population: Does the vaccine work or not in this combined population that includes a wide range of ages? But at the end of the day, when we look at the data, it is unlikely to be powered to look at subgroup-specific effects. In particular, we're interested in, how well does the vaccine work in older adults? How well does the vaccine work in people who have other comorbidities? And what does that mean for how we plan to strategically use a vaccine?

We know from vaccines for other diseases that sometimes they don't work as well in older adults. They don't generate the same immune response. So it is possible that we could end up with a vaccine that works well for younger people but doesn't work well for older adults.

In that case, it becomes increasingly important to have more than one vaccine. That impacts how we think about things like noninferiority and these comparisons, because there are so many in the pipeline. We want to find as many as we can that work.

Verghese: There are some estimates that one fifth of the population may decide not to take the vaccine. Assuming we have an effective vaccine, how does that play into our national strategy?

Dean: That's a major challenge. You can have a vaccine that works. But if people are not willing to take it, then that's going to represent a real barrier for achieving anything like herd immunity and for controlling the pandemic.

Social and behavioral research is always neglected in these types of pandemics, but it's incredibly important for the design and conduct of these trials. These trials have emphasized that they want a broad representation of different minority groups and different communities. But they've had challenges enrolling people.

Similarly, when we think about deploying a vaccine and who is going to consent to receive a vaccine, understanding how people make decisions and what messaging we should use is incredibly important. Also important is to have the process play out in a way that is transparent and trustworthy.

I think people — even people who would not describe themselves as vaccine skeptics or who are very pro-vaccine — see that there are concerns about different pressures on this process. People want to feel confident that the evidence is there to show the product is safe and effective.

Topol: I also wanted to mention the people who take the vaccine and think, I'm good to go, I'm immune, when that isn't necessarily the case. Can you comment on the fact that this is not sterilization immunity, that people may still potentially be infected, be carriers of the virus, and they will shed the virus? We will still be in a phase where masks and all the things we're doing now will be useful.

Dean: Absolutely. We don't exactly know the mechanism by which these vaccines will work. The ideal is that they will provide sterilizing immunity. But given some of the animal data we've seen, and given the fact that it's a respiratory virus, that may not be achievable. It is possible that a vaccine could at least reduce infectiousness, so you may be infected but you may not be as infectious to others. But we really just don't know.

We also may not know from the clinical trial data. The clinical trial data are going to tell us that the vaccine prevents disease. The public health goal is to reduce morbidity and mortality. The clinical trials will tell us if fewer people are symptomatic. It's possible for a vaccine to make people less symptomatic, but they could still be infectious to others. We're going to need more careful study on that. Many trials include infection as a secondary endpoint, so we'll have some understanding, but there is going to be uncertainty.

Also, it's not like we're going to have 80% coverage right away. It will take time to roll out, time to manufacture. There will be a long period when we need to be full force ahead on all the other things that we know work. We can't let up on all the other interventions just because a vaccine has promising data.

Topol: It's vital now to get to the other part of the exit strategy. We've covered that we're not going to be putting our masks away, and we're also not going to have unreasonable, short-term expectations of any vaccine, even when it's passed phase 3 and shown to suppress significant infections. Before all that, on the horizon we have rapid tests. Can you tell us where we stand with that and why it has promise?

Dean: Rapid testing is very promising. I really hope it can be approved and moved out soon. We have PCR testing, the nasal swabs, and those are very sensitive. They can detect even very low levels of virus. But it's a labor-intensive test. It has to go to the lab and someone has to run an assay — it's a process. We need to continue to look into ways to accelerate that process. But at the end of the day, it's never going to be as widespread as something like a test you could take at home or a rapid test.

Rapid tests give you a quick turnaround, and you can use them frequently. The goal is to be able to test people even without symptoms so that you can detect people during a period of high viral load. We know that viral load tends to peak a little bit before people develop symptoms, or around that time. You can't wait until someone develops symptoms to test them, which is largely what we've been doing so far.

You could imagine testing people all the time. Then you could spot people when they're very infectious. That allows you to refer those people to isolation and prevent those new chains of transmission. I see it not as a replacement for PCR testing but as a way to expand our reach and be able to find people during that period of peak infectiousness.

Verghese: I was reading a bit about your work with Zika and Ebola, and you used different strategies, such as ring vaccination. Is there room for some sort of innovation in the way we vaccinate, or the way we select people to vaccinate, that might be needed in a situation like this?

Dean: With ring vaccination, when you identify someone who is sick, you then identify their contacts and their contacts of contacts, which form a ring around this individual. You target these contacts for vaccinations. This is how smallpox was eradicated in the 1970s. It vaccinates people who are at highest risk, and it also creates this buffer by which the disease cannot escape beyond that population and that infected person.

A challenge here is that most of the COVID-19 vaccines, at least the early ones, are two-dose vaccines. They're not fast enough to be used in that reactive way. But we can still think about spatial units. Where are the parts of the country with the most transmission? And how do we want to target those parts of the country?

There's been a lot of discussion about vaccine allocation strategies and the optimal allocation. That's focused primarily on individual-level features, like someone's job, age group, the presence of comorbidities, or their social vulnerability. I'd like to see more discussion about the hot spots and how we are targeting hot spots in order to best control the virus.

Verghese: We haven't talked much about safety. What are your concerns regarding vaccine safety? What do we need to know to be able to make that declaration?

Dean: We need some level of follow-up that occurs shortly after someone receives the vaccine. Often, when vaccines have side effects, they're time-limited. They occur within a few weeks after the individual receives the vaccine. We've seen Guillain-Barré syndrome associated with a vaccine, and that's time-limited. So we need at least 1-2 months of follow-up in enough people after they receive their second dose.

The other safety concern we want to be able to confidently rule out is the possibility for enhanced disease and enhanced respiratory disease. The discussions I've heard indicate that this would be most likely to manifest when people are already infected; they would develop more severe disease.

That gets rolled into how we monitor the clinical trials. If we saw an excess of hospitalizations among those in the vaccinated arm, then that would signal a safety concern with the vaccine.

Then there's the longer-term safety data. We have less concern about long-term side effects. I haven't heard anyone posit real long-term side effects that we could expect.

With the dengue vaccine, there was another type of enhanced disease. Once the antibody response started to wane, then individuals were at greater risk for severe disease. That didn't appear until about 2-3 years after those people were vaccinated. But there isn't evidence that that's a concern with COVID-19. The challenge is that we're not going to have this longer-term safety data before we start using a vaccine in greater amounts.

Topol: One thing we don't know is the duration of protection the vaccine might give. I guess that could come from these phase 3 trials, if they keep the arms intact with the placebo group even after efficacy is established. Is that possible?

Dean: That's a question: Once you establish efficacy in an interim analysis, does everyone then receive vaccine? Some groups say yes, but others don't.

I've worked with the World Health Organization's Solidarity Trial that is testing COVID-19 treatments. The plan there is, as much as possible, to preserve the placebo arm until a vaccine is available, to preserve that time to be able to answer questions about long-term protection and long-term safety.

So that is an open question. We'll see how it plays out in practice. I think there will definitely be pressure to vaccinate people. But we have to be very clear about what we're losing in terms of the questions we can answer as a result.

Topol: The other thing about safety is the issue of possible immunopathologic reactions. Apparently, only 1% of participants in the vaccine trials are seropositive when they go in — meaning that very few people who have had an infection are enrolled — whereas in the United States, maybe 13%-14% of people have been infected, according to seropositivity, if you believe the numbers.

We're not going to be testing people when they get a vaccine. Are we going to have a problem where we will be vaccinating people who've had a prior infection, and now we have something we don't have our arms around?

Dean: Absolutely. We've seen, with the dengue vaccine, that the vaccine worked differently for people who were baseline seropositive vs people who were baseline seronegative. That impacts how these trials are designed. I believe one trial has excluded people who were seropositive at baseline.

But in general, what the trials in the United States have done is exclude people who are known to have had PCR-confirmed disease. But some of them are still seropositive. Ultimately, we want to make sure that the vaccine is safe and works well in everyone. That's why we need to make sure that the population that's in the trial is representative of the population we're going to use the vaccine in. And for the questions we can't answer in the trials, we're going to want to set up postlicensure studies and other types of studies to try to address some of these subgroup-specific questions.

Verghese: Do you think you have enough information, if you were offered one of the two vaccines that are being discussed, that you would say, "I'll take it"?

Dean: Right now?

Verghese: Yes.

Dean: No, because if I'm going to receive a vaccine, I want to at least contribute to understanding whether the vaccine works. I would participate within the context of a trial. Just deploying the vaccine without generating evidence — that doesn't make sense to me at this point. My main criteria would be: Are we gathering the evidence that we need to protect the wider population?

Topol: That brings me to the WHO, which has a global plan. You've worked with the WHO. But the United States has pulled out, diverted the funds, and isn't working in COVAX, which is this coordinated vaccine program. What is your sentiment about this go-it-alone strategy that we're seeing?

Dean: I think it's unfortunate because the reality is that we are all connected as a global population. That's how the pandemic spread so quickly. Our world is quite small when you see how connected we are, so a threat anywhere is a threat everywhere. If we're thinking just about the safety of Americans, then that also means that the world needs to be a safer place, because we're all connected. Of course, I work in public health because I'm concerned about people's health everywhere.

The WHO has a very challenging mission: to work with all of these diverse countries that have their own prerogatives and goals. But they do very important work, bringing people together to strategize about what we should be doing next. They've provided a real service with clear communications and transparency. And in a time when we haven't heard a lot from the CDC about digging through the evidence and what we should be doing, the daily press conferences at the WHO have been quite valuable.

Topol: They're doing such great things. For example, they had a meeting with 60 or more folks with long COVID, where the leaders of WHO actually sat with them to hear what it was about. You don't see anything like that going on here. I've been admiring their efforts, and I've gotten some insights through you.

The other thing I wanted to get your comments on: We are both on this advisory group for the COVID Tracking Project, which is a group of hundreds of volunteers who are trying to bring together data that otherwise might not be accessible. Can you comment on what that has done for the United States?

Dean: The COVID Tracking Project has been an incredible asset to all of us, because at a time when we all would have expected the federal government to step up and provide, as clean as possible, a standardized dataset for us to understand what was actually happening out in states, that wasn't available. And so a sort of vigilante group, a ragtag group of volunteers, came together, with The Atlantic magazine, and started collating the data.

It has been a real service because besides providing the dataset, it's provided understanding of the data. It has helped to set a benchmark for what we should be expecting from states. It has made states more accountable because it allows more people to lay eyes on the data. And that's what we need; we need more information, not less, so that we can make decisions.

Topol: You've been extraordinary in terms of recognized clear-eyed thinking and your posts on social media. If people aren't following Dr Dean on Twitter, they're really missing out. It's worth getting on Twitter just for that.

Tell us about how this has influenced your career — to be able to help provide guidance like this, to use social media to try to set things straight or organize your thoughts. What kind of impact has that had?

Dean: Thank you for your kind words. It has been a bizarre time to be someone who focuses on infectious disease research, and our fairly small community has gotten a lot more attention than usual.

I started using social media when I was on maternity leave and I could tweet while taking care of a baby. I realized that a lot of people wanted to learn and understand. I had this set of experiences from thinking about MERS, SARS, Ebola, and Zika, and vaccines, and all of this other stuff, and an ability to translate what other people were doing, and to provide some insights. I realized that there were a lot of people who wanted to learn, so I sort of leveraged my love of teaching.

Information is moving so quickly, and it's not going through the normal channels. It comes out in preprints and press releases, and people don't really know how to make sense of things.

I talk to a lot of reporters, but social media has allowed me to talk directly to people. I think that's been useful, and it's brought me to a lot of different experiences. I've been able to get to know a lot more people, within my field and in the wider biomedical field. That has been nice because it's very isolating to be at home, but in this way I've been able to make a lot more connections.

Verghese: It's been refreshing to hear from you and to follow your tweets. I only wish you were able to influence our national policy more. Clearly, we need this kind of clear thinking and wisdom to guide our vaccine strategy.

Topol: Let me echo that, Abraham. What also is dismaying to me is that Natalie's at the University of Florida, and Florida is one of the worst-hit states in the country. But they didn't turn to her, as best I know, to help guide them through what has been a tumultuous time. Hopefully that will change in the future. They're missing out on someone with extraordinary wisdom.

Natalie, we're so grateful to you for spending time with us today. We're going to continue to follow you closely because you are the antidote to the "infodemic." You are giving us good, straight information. Thanks for joining us today.

Eric J. Topol, MD, is one of the top 10 most cited researchers in medicine and frequently writes about technology in healthcare, including in his latest book, Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again.

Abraham Verghese, MD, is a critically acclaimed best-selling author and a physician with an international reputation for his focus on healing in an era when technology often overwhelms the human side of medicine.

Natalie Dean, PhD, is an assistant professor of biostatistics at the University of Florida in Gainesville. She specializes in emerging infectious diseases and vaccine study design. Follow her on Twitter .

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.