I grew up in the aftermath of the coup in Turkey, the one in 1980, following a generation that had a lot of grim experience with some of the worst conditions of repression—jails, detention, torture. They were not just older but often seemed impossibly distant from those of us who had not lived through any of the horrors they would sometimes hint at but rarely discuss openly but not really talk about. Their words often sounded like puzzles we could not make sense of, and their advice was cryptic. 

I remember a discussion about how one should never hope, but one also should never lose hope. I later learned that one of their defining experiences of the preceding generation was the detention period in the police headquarters. It was the worst phase—where the torture happened, and where people tried to endure and survive until they got transferred to the courts. They didn’t get justice in the courts, but they got relief from the worst. They’d see a judge and be sent to prison, which was certainly not a holiday but at least offered the relative safety of a ward with fellow prisoners. 

Apparently, one of the ways the torturers would try to break people during detention was to plant false hope—tomorrow you’ll be transferred, they’d promise—which, of course, didn’t happen, crushing people’s endurance in ways the horrific physical torture did not. The mind-games were more powerful, more cruel. But hope was also essential: knowing that it would end eventually was important and crucial.There were only so many days they could be kept before being transferred—the legal limit was thirty days weeks, I believe.While it wasn’t an ironclad guarantee, it meant something. Hope based on realism.

I have a hopeful piece out today. And there are a lot of reasons for hope this month. My piece

makes the case that people who’ve tried and failed to convince their elderly relatives to skip gatherings this holiday season should instead propose to their family members that they postpone it—just till March. 

The news from the vaccine front could hardly be better. It’s stunning, really. Both Moderna and Pfizer/BioNTech are reporting about 95% reduction in symptomatic disease. Preliminary Moderna data shows two-thirds reduction even in asymptomatic disease four weeks after the first dose—something that we don’t need for these vaccines to achieve to be spectacularly successful in ending this crisis, but is extra, if preliminary, good news on top of good news. Realistic hope is not a small thing.

The fight against the coronavirus has been called a “national marshmallow test” that we’re failing. In a famous study, children were left alone with a marshmallow for 15 minutes, and promised a second if they didn’t eat the first. Kids who were better at delaying gratification were found to be more successful later in life. At first, this correlation was explained as demonstrating the importance of willpower and executive function.

Later, a team of researchers set out to replicate this study and uncovered something profound. Once they adjusted for factors such as household income, mother’s education, and home environment at age 3, the effect disappeared. Further variations of the study showed that whether the children judged the promise to be reliable made a great difference in whether, and how long, they were willing to hold off for the reward. Indeed, access to a consistently well-stocked pantry makes it easier to believe those who say that a bigger reward awaits those who can resist eating the marshmallow right away. The precarity and instability of poverty encourage people to live in the moment, simply because the future is so uncertain. Willpower and grit are not merely personal characteristics, existing in a vacuum devoid of social reality. And, yes, hope works, but only when it is realistic and not an empty promise.

If we failed our national marshmallow test this summer and fall, perhaps that says something about how little reason the public was given for optimism. Hope can’t just be a slogan or a pep talk; it must be justified by facts, experiences, and trustworthy promises. And in fairness, until last month, it was less clear when and how this would all end.

But hope is justified today. Multiple vaccines are now in development.

I’ll have some more to write about this soon, on what I think the data do and don’t tell us, and what else we may be able to do in the short term. But I wanted to write a counter to my own hope.Because I think there is a plausible scenario in which the United States can snatch defeat from the jaws of victory, despite spectacular results in vaccine development: by not prioritizing vaccination correctly by risk.

You’ve all heard about it: this disease does not completely spare the young. Some healthy young people have a terrible time with it, some have had severe cases, and some have even died. There are long-term effects, yet their duration and prevalence are unknown. These are truths, important messages to get out there so that people don’t think they are invulnerable just because they are not in the highest-risk categories. These are all true.

But this simple fact is also true: the severity and death track one key variable more than anything else, and it’s age. The impact of age is not only huge, it’s exponential. As I wrote in my piece:

The risk profile of this disease is strikingly exponential: The risk of death for those ages 65 to 69 is a staggering two and a half times that of those just a decade younger. Those just a few years older, ages 75 to 79, face six times the risk of death compared with that same age group (ages 55 to 59). The steepness of this age curve really matters, because it means that protecting the most vulnerable groups with a highly efficacious vaccine will both quickly change our experience of the pandemic and relieve the strain on our hospitals.

It varies a little by country, but the numbers, everywhere, are staggering. In nearly all countries, almost all the deaths are from older people. In the United States, about 90% of the deaths are from people 55 and older. In Canada, it was about 95% of deaths from those above 65. In Italy, about 85 percent were 70 and older. And the gradations within those age groups are steep as well—hence the word, exponential. Unsurprisingly, severe disease and hospitalizations also track age.

When vaccinating under conditions of shortage, there are inescapable trade-offs. Obviously, vaccinating those most at risk is crucial. Transmission is always a consideration, so vaccinating people who either have a lot of contacts or have a lot of vulnerable contacts, is important. Often, as in this case, those groups do not overlap. There are also questions of equity: why should people who can work from home get the same priority as essential workers who have to work in person, and who take much higher risks? Shouldn’t those who have taken the most risk get priority?

In the end, though, we want to minimize human suffering and death. Overall, there seems to be a consensus that healthcare workers are going to be vaccinated first, along with long-term care residents, where a great majority of deaths have occurred. After that, the next question is whether to first vaccinate older people, starting with the oldest and working one’s way down the age range, or to start with essential workers, which are estimated around 80 million in this country.

It looks like the United States may first vaccinate essential workers—a category that will get defined somewhat subjectively, and according to the political power of these groups. A preliminary committee has already recommended vaccinating 80+ essential workers before vaccinating those 65 and older, and those with other conditions that put them at risk.The CDC will likely adopt this recommendation when they take up this issue on Sunday. After that, it will be up to each state to determine how they do this. Here’s what it may look like, with 85 million people being vaccinated ahead of those 65 and older.

But I’m already hearing that, for example, in Utah, a 30-year-old teacher may be vaccinated long before someone over 70 or even 80—even though the latter are at so much great risk if infected. In fact, it looks like teachers, police and food and agriculture workers will all precede adults over 65 and people with high-risk medical conditions.The predictable lobbying blitz has begun.

That is not what other countries rolling out the same vaccine are doing. For comparison, here’s the UK-wide vaccination prioritization, which sensibly ranks by risk, which corresponds to age.

Our proposed plan in the United States is also not what models show will be most effective in reducing death rates. Here’s a paper that models a whole range of scenarios: if a vaccine is rolled out before transmission begins, during an outbreak, if the vaccine blocks transmission or not, speed of rollout, etc. Under many different scenarios, vaccinating younger people reduces the rate of infections (they have higher exposure), but it’s almost invariably vaccinating the elderly that reduces deaths.

We also considered whether the rankings of prioritization strategies to minimize mortality would change if a vaccine were to block COVID-19 symptoms and mortality with 90% efficacy but with variable impact on SARS-CoV-2 infection and transmission. We found that direct vaccination of adults 60+ minimized mortality for all three Scenarios and for all vaccine supplies when up to 70% of transmission was blocked, and up to complete transmission blocking for most vaccine supply levels (Supplementary Text S1 and Supplementary Fig. S7).

Of course, none of this is to suggest that younger essential workers do not deserve to be vaccinated. Everyone deserves these vaccines and the protection they confer. But if we vaccinate those with exposure—or, frankly, those with more political power or who are looked at more sympathetically by decision-makers—rather than those at high risk, we are not going to get the potentially rapid relief hoped-for by my article. In writing that piece—postpone the holiday gathering—my hope was that by spring, we can at least have much lower death rates, and that we can decrease the strain on our hospitals (where most of those suffering from severe disease are the elderly). 

And while we are pretty confident about prevention of disease, and severe disease, from these vaccines, we aren’t equally confident on how “leaky” they will be for transmission: whether they will block all silent infections, or only some, and if so, what percent? Initial data is hopeful, for sure, but that’s yet another reason to prioritize the power of the vaccines we already know about: their incredible, stunning ability to prevent deaths, disease and severe disease, which disportionately stalks the elderly. 

These vaccines were developed under a timeline almost nobody thought was possible. They were based on a technology that had never been utilized at this scale before. The trials were so fast partly because of the crisis: the raging epidemic allowed for speedy comparisons between the vaccinated and the placebo groups. We’ve been delivered a near-miracle—based on research, creativity, investment and infrastructure. And yet, we are facing a terrible three months ahead, with overloaded hospitals and grim records being broken every day. 

We could choose to move as fast as we can to vaccinate those at most risk. That would follow the United Kingdom’s pattern. There, the oldest people get vaccinated first, and with other risk factors—the comorbidities—being considered as the groups get younger. (If we get to vaccinate someone younger, of course we should prioritize someone with diabetes compared with someone who is similar in all respects but doesn’t have diabetes, for example). 

Or, we could do what we appear to be planning to do: create complicated charts, and try to first vaccinate 80 million younger people as prioritized in a process open to influence by sensibilities and lobbying power. Then, in about twelve months, we can start writing those terrible papers comparing countries and states that chose to target for risk factors, where age stands above and beyond anything else in scale and scope, with those who did not. The findings will not be pretty, and the crushed optimism will have a cost which I dearly hope—that word again—that we can avoid.