Did the J&J pause boost confidence in vaccines?
On the day of the J&J pause Dr. Paul Offit, a member of the FDA’s Vaccine Advisory Committee, argued to a CNN panel that the pause should make people feel reassured:
Since then, a parade of government officials has tried to justify the CDC’s actions. Jerome Adams, the new Surgeon General, said on Twitter it was “the right thing to protect both people and vaccine confidence”. Jeff Zients, the White House Coronavirus Response Coordinator (essentially Deborah Birx’s successor) said this during a press conference on the day following the pause:
"We believe that by empowering Americans with data and facts, we will strengthen the public's trust in government, and increased their confidence in the vaccines”
In reality, the pause backfired in a terrible but entirely predictable way.
Most people accept that vaccines are the best solution to putting an end to the death and suffering caused by the ongoing COVID-19 pandemic. People know that J&J, as a one-shot vaccine that doesn’t require ultra cold cooling, is an important part of our vaccine arsenal. On the morning of the “pause”, when I searched for “J&J” I largely saw tweets from people expressing skepticism that the government would pause the vaccine over 6 severe clotting cases out of almost 7 million vaccinees. So people naturally believed that there must be additional data government officials were withholding - either data they were hiding or data they hadn’t collected yet due to holes in their monitoring system. It’s hard to dig up those tweets now, but I saw 5-6 independent tweets along these lines in just the space of a few minutes.
I realize that’s anecdotal. So, let’s look at the data.
Let’s first consider what happened during the AstraZeneca pause in Europe, which was also over a few-in-a-million blood clots. It is well known that humans suffer from cognitive biases which lead to irrational action in the face of small risks. Thus it’s not surprising that the number of people viewing the vaccine as unsafe grew significantly:
But notice that the numbers only grew in France and Germany, which implemented a temporary ban, while the numbers actually decreased slightly in the UK. This is quite surprising given the deluge of scary headlines in the news about the AstraZeneca blood clotting incidents. While scary headlines were inescapable everywhere, I suspect that UK coverage may have tended to emphasize the safety and life-saving nature of the vaccines while the EU coverage emphasized the need for precaution instad. For instance, this was an excellent BBC segment that explained clearly why Britons shouldn’t be worried and why the decision to stop the AstraZeneca rollout in many EU countries was irrational:
Of course, vaccine nationalism may also be part of the reason why Britons’ confidence in AstraZeneca’s vaccine stayed high.
Incidentally, the same polling firm, YouGov, also found that perceptions of the other vaccines didn’t change very much.
Now let’s look at some polling data from The Economist taken shortly after the J&J pause:
Not surprisingly, the figure on the right shows that more people now view J&J as “unsafe”.
We can also look at the figure on the left to get a clue how much the vaccine pause might effect people’s overall willingness to get vaccinated. The last data point in the fig on the left was collected after the J&J pause (they don’t say how many days after, though). It represents a 5% drop in those who plan to or have been vaccinated and a 2% rise in those who decided they won’t get it. Both are statistically insignificant but are in bad directions.
We can also look at this poll from Axios:
This poll is interesting because they had a category of people called "wait awhile & see" — these are people who are sort of on the fence. Among those, 62% said the news of the pause "negatively impacted their willingness to get vaccinated".
Now, here’s an example of an uninformative poll:
The relevant thing from a public health perspective is how many people went from wanting to get vaccinated to not wanting and vice versa as a result of the J&J pause. Obviously, this poll doesn’t tell us that. Asking people to take a side, such as “safe” or “unsafe” gives a much clearer window into what’s going on — if people switch from "safe" to "unsafe" that is a big deal. A change in confidence, however, may be small and ineffectual in terms of decision making. It’s rather sad this poll was touted by Vox in their article “Why the Johnson & Johnson vaccine pause was the right move”.
Finally, let’s look at vaccine uptake. It immediately reversed the day of the pause:
(Note this plot is smoothed with a 7 day rolling average). How much of the drop was from the J&J pause itself vs increased hesitancy towards other vaccines? Here’s the daily vaccines administered, unsmoothed:
As we can see, the number of J&J vaccines given starts dropping right after the pause (There must be reporting delays, and perhaps also a few places that gave out J&J vaccines anyway). We also see that the uptake of the Moderna and Pfizer went from rising to falling (trace the peaks of the oscillations).
On Twitter Daniel Bier has plotted the vaccine uptake data for different age groups, finding that nearly all age groups were affected to some degree or another but the drop-off in vaccinations was most pronounced among those between 18-64. That is troubling because research published in Science shows that the age group between 20-49 is most responsible for maintaining the spread of the virus (R0 > 1).
Overall this data clearly shows that the pause was the wrong decision. A better decision would have been to update the warning labels and do what the UK did - calmly explain that there is no reason for people to do be afraid and that the benefits of the vaccine, on both an individual and societal level, greatly outweigh the risks involved. Also they should explain, quite honestly, that many vaccines already in wide use, such as the MMR vaccine, carry 1-in-a-million risks of severe side effects. And to give people a sense of what that level of risk entails they might point out, as I’ve done before, that a 2-in-a-million risk is less than the risk of being struck by lightening each year or dying while driving a few hundred miles.
This saga continues a pattern we have same throughout the pandemic. Public officials have repeatedly made decisions largely on the basis of their own personal judgements of how people will react and how the reputations of themselves and their agency might be affected. This has resulted in a disastrous sequence of terrible decisions - people being told not to wear masks, the refusal to allow lab tests early on, the refusal to allow human challenge trials, the refusal to allow at-home testing, the delay in Pfizer’s EUA, the ban on the AstraZeneca vaccine, the refusal to do first doses first, and so on.
Every time public health officials have tried to impress us by playing “4D social chess” they’ve failed. Every time they’ve tried a noble lie, they’ve been quickly caught out. Their predictions on how the masses will respond have been consistently wrong and that is a symptom of rank elitism on full display.
Public health officials should be forced (via public audits, etc) to work within a structured decision making framework that maximizes QALYs saved using utilitarian calculations. The utilitarian calculus should be informed by scientifically informed and data-driven epidemiological simulations coupled with at least a rudimentary level scientific understanding of cognitive biases and human psychology. If systems that enforced structured rational decision making had been in place at the FDA and CDC many lives would have been saved.