by Sanjay Gupta
Bear in mind that a low chance of something happening—good or bad—does not mean that there is no chance. If the risk of a side effect developing is 1 person out of 100, that still means one person will experience that side effect. And it could be you.
When Johnson & Johnson temporarily paused its vaccine rollout due to the risk of blood clots, the media buzzed with confusing headlines. It presented a case study in how a new risk is evaluated in real time across a large population of people. In the United States, six women (out of nearly seven million vaccine recipients) developed a rare blood clotting disorder, vaccine-induced immune thrombotic thrombocytopenia (VITT), after receiving the J&J vaccine; two of them died from the condition. Naturally people panicked. These were women (maybe one man) younger than fifty years old, which made scientists wonder if gender and hormones had anything to do with it. The side effect appears to involve an immune response that differs from other types of clotting disorders and that predominantly affects women.6
The root of much of the anxiety may not have been the elevated risk of developing that side effect but rather simply not knowing how to put the risk into perspective. Here’s the paradoxical fact: The risk of having a clotting problem from a COVID infection is far greater than the risk of having a vaccine-associated clotting problem. Without question, the vaccine dramatically reduces the risk of any of the COVID-associated blood clots.
Blood clots are extremely common, affecting 900,000 Americans a year, according to the CDC.7 They kill an estimated 100,000 people annually. Clots in the brain are also common. About 795,000 people suffer strokes every year in the United States, according to the American Heart Association, and the group estimates 10 to 15 percent of these are in adults under the age of forty-five. Even the very specific type of clot associated with the vaccine, known as a cerebral venous sinus thrombosis, has a background rate of 5 people in 1 million in any given year.
Risk factors for ordinary blood clots include surgery, accidents, cancer treatments, hormonal birth control, smoking, and even sitting too long (prolonged sitting on long-haul flights, for instance, can substantially increase the risk for blood clots in vulnerable people). Although media reports popped up about people suffering more ordinary blood clots after having been vaccinated, it’s unlikely those were caused by the vaccine. But it’s hard for most people who aren’t trained in medicine to know the difference.8 Similarly, a person who unexpectedly has a massive heart attack (or gets hit by a car, for that matter) the day after getting a vaccine may think the event is linked to or even caused by the shot. We know that this is not the case.
The chances, or absolute risk, of experiencing clotting complications as a result of either the J&J vaccine or the AstraZeneca one that is based on similar technology is 1 in 1 million (about the same odds, relatively speaking, as dying in a plane crash but much lower odds than going to the ER with a pogo stick–related injury, the latter of which is entirely preventable but affects 1 in 115,300 Americans). The risk of a blood clot for a young woman taking the birth control pill is about the same as the risk of a person being struck by lightning.
I should also add that if you get a flu shot, there’s a 1 in 1 million chance of developing Guillain-Barré syndrome, a rare disease that can cause paralysis. Another way to look at these relative numbers is the following: If 1 million people contract COVID, roughly 5,000 will die based on current data. If 1 million people receive the J&J vaccine, maybe one person will develop this specific type of blood clot in the brain, which is treatable if diagnosed early. Which side of the odds do you want to be on?
For me, it was easy. I picked the vaccine, and so did my wife, who was in the age group of women at risk. Part of her decision was also informed by the fact that we know how to treat adverse reactions to these vaccines. These conditions, from VITT to allergic reactions, are manageable and curable. Doctors know what to look for and treat accordingly.
Let’s Play Risk9
Common Risks
The Chances
Dying in a road traffic accident over fifty years of driving
1 in 85
Needing emergency treatment in the next year from an injury by a can, glass bottle, or jar
1 in 1,000
Needing emergency treatment in the next year from an injury by a bed, mattress, or pillow
1 in 2,000
Dying in any accident at home in the next year
1 in 7,100
Being hit in your home by a crashing airplane
1 in 250,000
Drowning in the bath in the next year
1 in 685,000
Throughout 2020, we didn’t have enough information to put COVID into risk context. But now we have more data to make reasonable assessments about an individual’s risk for contracting the virus and how that person will fare in the course of the disease based on a few variables such as age, health status, and access to care. It is true that the nature of a contagious disease means you need to not only account for your own risk but also the risk you may pose to others. Fortunately, there are ways to mitigate risk. It depends on three important factors:
What you do. What kinds of activities are you engaged in, and who is around you? What is the likelihood you’re going to breathe someone else’s air or others will breathe yours?
Where you are. Your location in the world locally determines your risk for exposure (e.g., indoors versus outdoors, in areas with high versus low COVID transmission, being where the majority of people around you are vaccinated or not.)
What you bring. Do you have any personal risk factors like preexisting conditions that could complicate a COVID infection?
Rewriting Risk: Avoid the Traps
It is important to be aware that as your brain calculates risk throughout the day, it goes to a default emotional response based on limited or overly biased information. Here are some competing interests that could be hampering your ability to assess risk fairly:
• It’s not going to happen to me. This is called optimism bias, and it’s one of the most basic, well-established principles in social psychology. People with optimism bias think their own risk is less than other people’s risk. This type of bias is more prominent in individualistic, mostly Western societies like ours, where we prioritize personal choice and the rights of individuals (compared to collectivistic cultures, where the focus is on group goals and what’s best for the group). And it explains why you’ll choose to eat a cheeseburger and fries over fish and steamed broccoli. It’s not that you don’t believe heart disease is more associated with meals higher in saturated fat. You understand that, but you also think the risk is just higher for other people.
• I’m in total control so I’ll be okay. When we feel in control, even if it’s a false sense of control, we’re less likely to be worried. Driving a car seems safer than being a passenger on a plane, but the data make clear that is not the case. Driving long distances and making pit stops can entail more potential exposures than flying on a plane point to point. Similarly, when we follow public health advice and wear masks, wash our hands, and practice social distancing, our perceived risk of contracting the virus is lower and can make us act in a more cavalier way.
• No one knows, so why should I worry? Mixed messaging from the beginning of the pandemic from public health experts and other leaders has not helped create a unified front about the dangers of COVID. The messages around masks alone were divisive and confusing, giving some people permission to lessen their sense of true risk.
• My social circles have all stayed healthy even though they haven’t followed the public health measures. This fallacy reminds me of my smoking patients who tell me they are not worried because they know someone who smoked their entire life and never developed lung cancer. They search for messages that reinforce what they want to hear. This happened even more acutely during COVID: The anxiety and strain imposed by the disease led people to huddle inside their like-minded groups and develop their own groupthink. The group, in many cases, bec
ame their identity.
• I read that it’s safe to eat indoors if the tables are spaced well enough apart. This is another example of confirmation bias. You wanted to confirm a hope that it’s fine to dine inside. So you search for the answer you want to find by typing in “safe dining indoors during COVID” on the Internet rather than “dangers to dining inside during COVID.” Each of those searches will turn up radically different sources of information, and you will gravitate toward the ones that will assure you it’s okay because they’re aligned with what you wish the outcome to be.
• I’ve been going to the grocery store and running errands for months now; it’s totally safe. Exposure therapy is real: The first trip outside your home in public following a lockdown may have stressed you out and made you anxious. But after braving public settings a few times and not getting sick, the outings lose their risky edge, and it’s natural to let your guard down. Those subsequent trips seem less scary, but your risk may have been increasing as the actual COVID numbers were climbing in your area. The risk tied to running errands does not remain the same—it’s dynamic with the ever-changing rates of community spread. Over the past year, I have spoken to hundreds of patients who became ill from COVID. Invariably they always said how surprised they were to actually have contracted the illness. Their minds had rewritten the risk even as the worrying evidence was there to see.
Risk Going Forward
Now, more than a year into the pandemic, we know exponentially more about this disease than at the beginning. But it is still a small fraction of what we need to learn to assess risk. It is still unclear, for example, why some people breeze through their illness while someone who is similar in age and health background will be undone by it. “I want to find out how it could possibly be that the same virus that’s killed more than half a million people in this country is a virus in which more than half the people don’t ever get any symptoms,” Fauci told me in spring 2021. It can be wildly different even for identical twins, like Kelly and Kimberly Standard, who also lived together. I spoke to the thirty-five-year-old sisters about what happened to them in spring 2020.10
After experiencing fever and shortness of breath that hadn’t gotten better after a few days, both went to the emergency room and were diagnosed with COVID. Kelly said she had a bad feeling about the situation. “I’ve got high blood pressure. I’m diabetic. I have breathing problems—I’m asthmatic—and I think this virus is really going to affect me. I was thinking, It’s going to get worse,” she told me. Like her twin, Kimberly had similar medical conditions but said she felt the “complete opposite”: she wasn’t really worried. “In my mind, I’m thinking, Okay, let’s get this out of the way and go home.”
The Standard sisters were admitted to Ascension Providence Rochester Hospital in Michigan on the same day, and after that, everything changed. Kelly, who had the bad premonition, got better with treatment and was discharged; Kimberly became much worse. She was airlifted by helicopter to a different hospital where she eventually wound up on a type of life support called ECMO—extracorporeal membrane oxygenation. It’s a machine that pumps and oxygenates a patient’s blood outside the body. She spent about a month connected to tubes and machines, in and out of consciousness, battling for her life.
This vast difference in how Kelly and Kimberly reacted to the coronavirus surprised the twins and their doctors. There is still a significant randomness to how COVID affects certain individuals, even twins, and that makes calculating risk going forward all the more difficult.
The 5 Percent Rule
As a general rule and reminder, you want to be in a setting where the positivity rate—the percentage of people who test positive for the virus—is below 5 percent. Positivity rates are a confusing concept for people, but here is a way to think about it. If you are fishing with a net and you bring up lots of fish, that probably means there are a lot more fish down there you are still missing: a high positivity rate. If you only bring up a few fish, a low positivity rate: that means you are probably catching most of them. But here’s the thing: Sometimes you cannot know the positivity rate in your exact location. Or by the time you do, it’s too late and you’ve been exposed. With that in mind, it is reassuring to know that regular testing could become the norm for this pandemic and future ones. Many at-home rapid test kits are coming on the market that will help us monitor where the virus is lurking and potentially spreading. This will help us get ahead of future outbreaks and keep infected people isolated.
Keep in mind too that community immunity is not necessarily permanent. It is based on the contagiousness of the virus at any given time. With COVID, the threshold for adequate immunity may be closer to 70 percent in the summer months when heat and humidity slow viral transmission. It may then pop back up to 80 percent in the cooler and dryer winter months, which is why it is so important to achieve high vaccination rates even as cases are dropping. As the threat of the virus ebbs and flows over the years, so too will risk factors.
Dynamic Viruses Demand Dynamic Responses: Place Your Bet
With a germ like smallpox, a single vaccination conferred lifetime immunity. Smallpox is a double-stranded DNA virus that lacks a known animal reservoir other than humans. Compare that to COVID, which is a single-stranded RNA virus with a higher mutation rate and multiple animal hosts. Average mutation rates in RNA viruses are estimated to be about a hundred times higher than those for DNA viruses and up to a million times higher than their hosts’. Put simply, we cannot mutate fast enough to gain immunity against COVID and so we must accommodate it as an ever-present threat by constantly evaluating risk within the context of our lives and taking necessary precautions to prevent exposure.
One provocative way to put risk into perspective is to think like a poker player. When I spoke with psychologist, champion poker player, and author of The Biggest Bluff Maria Konnikova, she helped me understand how to think through the probable risks associated with different decisions as a serious poker player. It was an astonishing lesson for me. Maria first got into the world of poker out of a curiosity for trying to disentangle skill from chance. She wanted to find out: Where are the limits of control? Where are the limits of what we can and can’t do? And where does variance enter into it? When does pure luck enter into it? How do you learn to tell the difference? We both agreed that life is a game of incomplete information; you never know everything, so the question then becomes: Do you try and always complete fully gathering the information first, or do you try and get really good at making decisions with the information that you have?
“I think you need to get comfortable with uncertainty,” Maria told me, “and with the fact that any decision is going to be inherently probabilistic.” She’s right. As she reiterated to me, there’s no such thing as certainty in anything in life. No such thing as 100 percent. Everything is probabilistic. If you get up to 98 percent, you’re ecstatic. But two percent is a lot. One percent is a lot. All of those tiny percentages are actually huge when it comes to talking about billions of people and billions of outcomes. And as the stakes go up, those tiny percentages have to mean more. So the best you can do is to make the best decision you can with the information you have, knowing that it’s never going to be perfect. “I think that quest for perfection can actually hinder us more than it can help us,” Maria said. She also brought up a good point: We process risk through experience, but our experiences are not necessarily representative of actual statistical risk—they are skewed. So sometimes we will overestimate tiny risks and sometimes we’ll underestimate risks that are actually much bigger simply because our personal experience has skewed us too far in one direction or another. Our personal biases get in the way. So, how can one correct for this? Enter a game of poker where, as in everyday life, there are consequences for keeping it too safe… and there are consequences for taking too much risk.
As Maria described it, poker is all about adjusting your strategy based on the circumstances. Sometimes it’s going to pay to take more risks
than you normally would, and sometimes it will pay to be more conservative. And the same strategy doesn’t always work.
“If you’re playing a game, especially tournament poker, where the stakes keep going up, you’re going to go broke if you take too many risks. You’re going to bleed chips. On the other hand, if you haven’t been betting much, then when you do have a strong hand and you’re ready to take the risk, everyone’s going to fold. Everyone’s going to be scared of you because all of a sudden you got very aggressive. If people aren’t idiots, they’ll realize, Okay, you know, this person has a strong hand. I need to get out of the way. And so you’re not going to be able to maximize money even when you have good cards.… You need to find that magic middle.”
For Maria, poker represents a risk exercise based on incomplete information, your own risk tolerance, and understanding how others will react to you in a dynamic situation. It is not a perfect metaphor, though, because the politicization of the pandemic in America was a wild card. It made it uniquely different from a game of poker, as people were more likely to identify with certain data that met their political party’s thinking, even if it made it more likely you would lose money, or your own life. In a pandemic like this one, evaluating risk has also become a question of identity—whose tribe and way of thinking you subscribe to.
But politics aside, Maria’s idea of placing a bet on a decision is a great way to assess your sense of risk. Put simply, the best way to understand uncertainty is to bet on it. How much are you willing to bet you’re right about a certain risk, or wrong? This is a strategy borrowed from Immanuel Kant, who once used an apt analogy in the medical world where people tend to have a false sense of confidence in their doctor. Maria recalled the thought experiment well for me: Imagine you go to a doctor and the doctor looks at you and gives you a diagnosis and you leave. Now what if you actually stopped for a second and forced that doctor to put money on the diagnosis? How much would the doctor be willing to bet that the diagnosis is correct? $10? $100? $1,000? $10,000? $100,000? His marriage? His happiness? His life? Where’s the line? “Something like that is an incredibly powerful corrective to overconfidence and to false certainty,” Maria said. “All of a sudden, you’re forced to actually ask yourself, Wait, well, what’s my basis for making this?”