F. Perry Wilson | Vox

A doctor on 9 things that could go wrong with the new vaccines

2020-12-14T11:56:47-05:00

Nurses check the dose amount of the Pfizer/BioNTech Covid-19 vaccine before a husband and wife receive the injection in Essex, England. | Victoria Jones/PA Images via Getty Images

Think of the achievement. Within a month of the emergence of SARS-CoV-2 — the coronavirus that causes Covid-19 — its genome was sequenced. Three months later, the first vaccine candidates were being injected into human volunteers in clinical trials.

Now, less than 12 months after the first case was identified in Wuhan, China, the US is slated to initiate the largest mass-vaccination program in its history. Few achievements in modern science rival the speed and the audacity of the coronavirus vaccine program. With the Pfizer/BioNTech vaccine’s imminent emergency use authorization from the US Food and Drug Administration, it feels as though the long dark of the pandemic — which has claimed over 283,000 American lives and more than 1.5 million worldwide — will soon be relegated to the litany of global tragedies, a thing of the past.

As a physician, clinical researcher, and epidemiologist, I am thrilled with the vaccine data so far. The 95 percent efficacy of the Pfizer/BioNTech and Moderna mRNA vaccines is unprecedented and better than any of us hoped for.

But we need to be careful. We need to temper our enthusiasm with the acknowledgment that the vaccine is a weapon we may not be fully prepared to wield.

A lot can still go wrong.

I lay out my worries here not as a wet blanket, but because I am a worrier. And, like all the worriers out there, one of the reasons I worry is to ensure the things I worry about don’t actually come to pass.

By worrying together, we can prevent much of this from happening. I’m providing my worries in a convenient list format, from low probability to high.

1) Unexpected long-term side effects (probability: low)

Though mRNA vaccines have never been used in a broad-scale vaccination effort before, in theory, there’s not a ton that can go wrong here.

The mRNA molecule is incredibly unstable — it’s so easy to break down it needs to be transported in super-cold conditions. It has no ability to integrate into DNA, so there’s no risk of sci-fi human/coronavirus mutants cropping up. Some scientists have raised concerns that an immune reaction to RNA could lead to some autoimmune diseases (like lupus) down the road, but extensive human testing has not shown this yet.

2) There won’t be enough vaccine for everyone (probability: low)

We’re likely in good shape here, if we define vaccine as “that stuff that goes in a vial and eventually into your arm.” One of the key advantages of the mRNA vaccines is that they are pretty easy to scale up. In fact, you can make about 1 million doses in a bioreactor the size of a Coke bottle.

But we don’t have to depend on the mRNA vaccines alone. AstraZeneca’s vaccine (which contains a bit of coronavirus genetic material wrapped in an adenovirus shell) has suffered some setbacks recently but is still likely to join the fight in the next few months. Johnson & Johnson’s vaccine — notable for being a one-dose regimen — may receive an emergency use authorization in early 2021.

And the pipeline is full: There are currently 13 vaccines in phase 3 trials (not including the Pfizer/BioNTech vaccine), 17 in phase 2, and a slew of others still in the early stages of testing. Many of these vaccine candidates target the coronavirus spike protein — the same target that has led to high efficacy rates in the frontrunners. If these trials can recruit participants quickly, our armamentarium will increase dramatically.

But there is one concern: The existence of effective vaccines (like Pfizer/BioNTech’s and Moderna’s) may dampen recruitment in ongoing trials. Enrollment in a trial means you have a typically 50 percent chance of receiving a placebo instead of a vaccine. Will individuals continue to volunteer when an effective vaccine may soon be on the market? We’ll know soon enough.

3) Vaccination becomes politicized (probability: low)

Am I too Pollyannaish on this one? I can’t imagine that Republicans, who are desperate to reopen society at any cost, would look the vaccine gift horse in the mouth. Nevertheless, the current leader of the Republican Party has been a “vaccine skeptic.”

That said, the vaccine may prove to be the most broadly popular achievement of Donald Trump’s presidency. And I encourage him to take credit — his crowing about the vaccine may increase vaccination rates among his supporters, many of whom are increasingly skeptical of science and have “low social trust.”

4) There won’t be enough vaccine supplies (probability: medium)

This is more of a problem. Vaccines require glass vials, needles, alcohol swabs, and — in the case of the Pfizer/BioNTech vaccine — an amazing amount of dry ice. Pfizer has actually cut its vaccine delivery estimates in half because of unspecified shortages of these products. These products are generally not produced by the vaccine manufacturer and need to be sourced from other companies.

Despite the success of Operation Warp Speed at accelerating vaccine development, the federal government has resisted calls to invoke the Defense Production Act to increase the supply of these critical components of the vaccination program.

So we face the possibility that we may have plenty of vaccine in freezers but not enough needles to get them into enough arms. This would slow the vaccine rollout, prolonging the pandemic despite approved vaccine candidates.

5) People won’t get both doses (probability: medium)

Multi-dose vaccines are nothing new — measles, mumps, rubella (MMR), rotavirus, and tetanus are all multi-dose regimens. But those are routine childhood vaccines, and parents tend to be a bit more obsessive about their children’s health than they are about their own.

Parallels to seasonal flu vaccine aren’t helping either. Getting a yearly flu shot is easy because it’s only one shot. And, to be frank, we’re not even that good at getting our yearly flu vaccine (only 45 percent of US adults got their flu vaccine in the 2018-2019 flu season).

How many of us will remember to go back three weeks later for the booster Covid-19 shot? This has a few real consequences. First, people may feel protected when they aren’t — leading to behavior (crowded post-vaccine dinner parties?) that will spread the virus even faster.

But there’s a bigger concern here. Several virologists have suggested that inadequate vaccinations may allow the virus to develop “vaccine resistance.” The idea here is that a partially vaccinated person may get a low-grade infection, and selection pressures within that individual will favor viruses that can evade the vaccine-induced immune response.

I asked Yale immunologist Akiko Iwasaki about the possibility: “It is conceivable that people who have only one shot of the vaccine … can become infected, resulting in an escape mutant,” she said. “This is purely hypothetical at this time, as we have yet to see such escape mutants arise from naturally infected people who have suboptimal immunity.”

So what do we do to get people their second shot? Here are some ideas, from Vox’s Dylan Scott.

6) Doctors will bend the truth to help their patients get a vaccine faster (probability: medium)

Until vaccines are ubiquitous, we’ll need to prioritize who gets them first and who has to wait. The Centers for Disease Control and Prevention has released its initial guidance, focusing on health care workers and residents of long-term care facilities for the first phase of the rollout.

Phase 2 is where it gets tricky. We don’t have great clarity yet, but it is likely that the vaccine will be reserved for those with preexisting conditions (like older age, diabetes, and maybe even obesity). How will your local CVS know that you have these comorbidities? Your doctor will most likely have to vouch for you.

This will create some perverse incentives for doctors like me, who often think more about the benefit to individual patients than we do to society at large. Should I check that blood pressure one more time to see if I can make a hypertension diagnosis? Should I check the “obesity” mark on the vaccine form even if the patient is just overweight? And lest you think doctors are immune from this type of behavior, let me remind you how often we give our patients antibiotics for no rational reason — 30 percent in the largest study to date.

We want to keep our patients happy. That’s not always the right thing to do. There’s certainly an unfairness issue, but the bigger problem with doctors upgrading people to “high risk” is the inappropriate allocation of scarce vaccine resources to those who would likely do okay without it. While vaccines are scarce, it’s critical that we vaccinate the actual highest-risk groups, not the individuals that doctors can make look high risk.

7) Vaccines will exacerbate inequality in the health care system (probability: high)

This is really a corollary to No. 3, but something that keeps me up at night. There are currently 80 million people in the US with no regular access to doctor’s care, many of whom have significant comorbidities that no one is documenting. These are predominantly people of color and of lower socioeconomic status. These are also the people who have suffered most during the Covid-19 pandemic.

In other words, they are the people who would most benefit from the vaccine. And they may be left behind.

To prevent this, we need targeted vaccination programs in low-income and underresourced communities. We also need to waive comorbidity restrictions among those without access to quality health care. California’s proposal to consider “historical injustice” to vaccine allocation is not far off the mark.

8) A false sense of security develops (probability: high)

Ninety-five percent efficacy is nothing to sneeze at. At that level of efficacy, even if people loosen up, go out to dinner, and wear masks less after they get vaccinated, we should still see a dramatic decline in infections.

The problem with the false sense of security is not societal, it’s individual. When a vaccine is 95 percent effective, everyone who gets it assumes they are in the 95 percent. No one thinks they are in the 5 percent, but 5 out of every 100 people are. If those five people stop engaging in behaviors like social distancing and mask-wearing before the pandemic ends, they may still suffer the gravest consequences of Covid-19.

9) Anti-vaxxers amplify and misrepresent side effects (probability: almost certain)

This is already happening. The real worry is how much it will affect the broad-scale vaccination effort. We need to get about 70 percent of the population vaccinated (or infected with Covid-19, which would be ethically wrong) for herd immunity to end the pandemic. That’s a high bar, and the power of social media to amplify false or misinterpreted messages and spread them far and wide is huge. I’m not worried about the “microchip tracking devices” nonsense. I’m worried about anecdotes.

I remember a story Paul Offit, a vaccine expert at the University of Pennsylvania and member of the FDA vaccine advisory committee, told me once when I was in residency. He was about to vaccinate a child with the MMR vaccine. Five minutes before he gave the vaccine, the child had the first seizure of his life. Can you imagine what would have happened if that seizure occurred five minutes after he gave the vaccine?

We’re going to vaccinate hundreds of millions of people. Someone is going to have a seizure after they get the vaccine. Someone is going to have a heart attack. Someone will get into a car accident, and someone will die by suicide. These stories will burn through social media like wildfire. And remember that anecdote and evidence are not the same thing.

Are there other things to worry about? Absolutely. The thing that derails the whole effort may be something none of us has thought of yet. It will be important to stay flexible, stay optimistic, confront the challenges as they come, and keep hope alive so we can keep each other alive.

F. Perry Wilson, MD MSCE is an associate professor of medicine at the Yale School of Medicine and director of Yale’s Clinical and Translational Research Accelerator. He writes a weekly column on Medscape.com and is the creator of the free online course “Understanding Medical Research: Your Facebook Friend Is Wrong.”

What if the Covid-19 vaccine only works half the time?

2020-11-16T16:16:50-05:00

A health worker gets a vaccine as part of a drill to prepare for the arrival of a Covid-19 vaccine at a health clinic in Depok, West Java, Indonesia. | Jefta Images/Barcroft Media via Getty Images

In the not-so-distant future — we’ll probably know who won the 2020 election, but memories of the weirdest Thanksgiving ever will not have faded — you may be faced with an incredibly important choice. Should you get the Covid-19 vaccine?

It’s been less than a year since the discovery of SARS-CoV-2, the virus that causes the new disease, and we have not just one, but multiple vaccine candidates that may in time truly shut the door on the pandemic.

Though nothing is certain, it’s very likely that some vaccine will be approved soon — and widely available by the end of 2021.

Which leads us to ask an important question: What if that vaccine is not very good?

The paradox of the mediocre vaccine

What do I mean by not very good? 50 percent efficacy. Food and Drug Administration commissioner Stephen Hahn has pledged that the FDA will not grant approval for a vaccine that has less than 50 percent efficacy, so this seems like a reasonable floor.

To give a sense of what vaccine efficacy really means, let’s imagine ourselves in the board rooms and meeting spaces at vaccine manufacturers Pfizer, Moderna, and their ilk as they wait for the results to come in. Each of these companies is running large, randomized clinical trials of 30,000 to 60,000 individuals who have not yet contracted Covid-19 in areas where the virus is prevalent. Half get the vaccine; half get a placebo shot. The companies then wait to see who gets sick.

The best-case scenario for the manufacturer? Zero cases in the vaccinated group and something well north of zero cases in the placebo group. With, say, 150 cases in the placebo group and zero in the vaccine group, they would be able to claim that their vaccine was nearly 100 percent effective. (To be clear, no vaccine has ever been 100 percent effective.)

If there are 150 cases in the placebo group and 75 cases in the vaccine group, now they could say that their vaccine is 50 percent effective. Not as impressive, clearly, but something is still happening. Both Pfizer and Moderna planned on presenting early results after 150 infections had occurred in their study population.

But here’s the problem. The less effective a vaccine is, the more of us need to get it to end the pandemic.

The math is pretty straightforward. Let’s say that, on average, an individual infected with Covid-19 infects two more people (the now familiar R0). Unchecked, that leads to a huge growth in infections.

To end the spread, we need to make sure that each infected person infects less than one additional person. That can be achieved through mask-wearing, social distancing, canceling large gatherings, you know the drill. But it’s not ideal — we want to lift all the restrictions. We need one out of every two people to be immune.

And there’s two ways to become immune — through infection or vaccination. (Caveat: The mechanics and durability of Covid-19 immunity are still being studied).

If there is a 100 percent effective vaccine, and one out of every two people need to be immune, it means that 50 percent of the population would need to be vaccinated to end the pandemic.

But what if we had a 50 percent effective vaccine? In that case, one out of every two people who is vaccinated won’t be protected. To end the pandemic under this scenario, 100 percent of the population would need to be vaccinated.

The worse the vaccine is, the more people need to get it to get us out of this mess. The less effective a vaccine is, the less likely people will choose to get it.

A new survey in JAMA Network Open evaluating factors predictive of whether people would accept the vaccine found that vaccine efficacy was the No. 1 determinant. People were, on average, 16 percent more likely to accept a vaccine with 90 percent efficacy compared to one with 50 percent efficacy.

Get the vaccine anyway

It’s tempting to look at the math of a mediocre vaccine and give up, but a mediocre vaccine can still end the pandemic — just not on its own.

The key is the R0. My example assumed that the average infected person infects two more people, but it doesn’t have to be that way. In fact, due to the policies enacted to stem the tide of the virus, the effective R-value (often called Rt) is probably lower than 2 — maybe around 1.2.

If that’s the case, you’d only need to get a 50 percent effective vaccine to 33 percent of the population to end the pandemic, provided we all continue to wear masks until it’s over.

Most polls suggest that around 70 percent of Americans will get a Covid-19 vaccine when it is available. A 50 percent effective vaccine may have different efficacy in different groups — if a vaccine is highly effective in young people but ineffective in older people, it would complicate matters quite a bit, as this new perspective in the journal Science notes.

What about side effects?

Efficacy is not the only thing to think about when you are lining up to receive a vaccine. The side effects are critical, too.

Researchers categorize adverse events as serious or not depending on how severely they affect the patient. Redness and swelling at the injection site? Not serious. Anaphylaxis that lands you in the hospital? Serious.

There’s no need to be too worried about non-serious adverse events. Every vaccine will have them: fevers, injection site reactions, malaise. It’s the serious adverse events that we need to worry about, and fortunately these are usually rare.

Of course, it is that rareness that leads to uncertainty. Let’s say I’m Moderna and I enroll 30,000 people in a vaccine trial and not a single one develops anaphylaxis, a life-threatening allergic reaction. You might think that implies that the anaphylaxis rate for this vaccine is zero percent, or at least less than 1 in 30,0000. You’d be wrong.

The vicissitudes of chance become clearer if we make the numbers smaller. Imagine this trial only enrolled 10 patients, and, again, none developed anaphylaxis. It would be wrong for me to conclude that anaphylaxis can never happen with this vaccine.

All I can say with absolute certainty is that the anaphylaxis rate is less than 100 percent. But it could be 99 percent. I may have just been lucky enough to enroll 10 people who don’t get anaphylaxis from my vaccine.

We need some way to estimate a plausible side-effect risk, given the observed data. This can be done with some simple statistics.

We can ask what is the plausible range for the anaphylaxis rate given that I saw it zero times out of 10. We’ll define plausible by saying that, whatever the real rate is, I would have a greater than 2.5 percent chance of not seeing it at all in my 10 patient sample (this is a standard rate of plausibility for this type of thing). The answer: 0 to 31 percent.

In other words, if the true anaphylaxis rate is 50 percent, it would be unusual to enroll 10 people and not see a single case. But if the true anaphylaxis rate is 30 percent, I’d have a shot to pitch a no-hitter (I’d get data like mine around 2.5 percent of the time).

That’s a huge range, but fortunately the ongoing vaccine trials enrolled many more than 10 people. How certain can we be about side effects? For serious adverse events that don’t occur at all in the trial, we can be confident that the true rate is less than one in 10,000. That is a really small number, but not necessarily reassuring when we’re talking about vaccinating 350 million Americans (or nearly 8 billion people).

Given that, you may think it’s appropriate to take a wait-and-see approach. Though there is a bit of moral hazard (if we all act this way, we’ll never learn anything and the pandemic will march on), it’s not an untenable position.

In fact, since it’s extremely unlikely that mass vaccination will be available to everyone at the same time, those at highest risk of Covid-19 (including health care workers like me) will be prioritized, take on the unknown risks, and be followed closely for side effects. By the time the vaccine is ready for everyone, we should know much more.

Me versus us

This pandemic has brought to light a particular weakness in American society: a seemingly irreconcilable tension between a perception of individual liberty and collective sacrifice. Nowhere has this been clearer than in the case of face masks, which carry zero risk, provide modest individual protection, and have a strong societal benefit.

Mediocre vaccines are more like face masks than miracles. The real benefit of getting vaccinated may be less to the individual and more to society. Sure, a 50 percent effective vaccine buys you some peace of mind. You’ve cut your risk of infection in half, and (since vaccine effectiveness counts all infections the same, regardless of how severe they are) even if you become infected you may get less sick. But in the end, if all we have is a mediocre vaccine, the goal needs to shift from providing individual protection to ending the pandemic.

That means we need to start convincing people to get vaccinated now. We need to prepare the public for a less-than-ideal scenario — a mediocre vaccine — and show why it is important to get vaccinated anyway. A mediocre vaccine won’t necessarily save you, but if enough of us get it, it can save us.

F. Perry Wilson, MD MSCE (@methodsmanmd) is an associate professor of medicine at the Yale School of Medicine and director of Yale’s Clinical and Translational Research Accelerator. He writes a weekly column on Medscape.com and is the creator of the free online course “Understanding Medical Research: Your Facebook Friend is Wrong.”