The Muddled Speech of Numbers

Blood clots, COVID-19 vaccines, and statistical risk

Earlier this week, the CDC paused the roll-out of the Johnson & Johnson COVID-19 vaccination after 6 women experienced serious blood clots. Their caution has merit, given that the FDA has been approving vaccinations in advance of the typical large-scale evaluations because speed is seen as so crucial. Reasonably, there is a desire to know more about these blood clots before more might appear. Yet, there was also sheer frustration from many in the medical community because the choice to pause the roll-out suggested that there was a serious issue, that the vaccine was dangerous. In a context in which vaccine hesitancy is likely to undermine herd immunity, any suggestion that the vaccine might have consequences can be twisted and contorted. 

Across many mailing lists and Twitter streams, I kept seeing data points trying to ground the seriousness of the blood clots in the J&J vaccine. Most referenced the frequency of blood clots that women experience while taking the birth control pill, roughly 1/1000. People also highlighted how common blood clots are for those who are in the throes of COVID-19. These were meant to highlight just how rare and statistically insignificant blood clots are when taking the J&J vaccine. 

Yet, as these attempts to ground the conversation unfolded, a different kind of outrage formed. A handful of people highlighted women they knew who had died of blood clots most likely related to birth control. Many more women who took hormonal birth control expressed frustration that they had no idea that they were at increased risk of a blood clot. Sure, it's part of the fine print of that printout you get from CVS when picking up your pill, but this wasn't something doctors emphasized. Unlike the J&J vaccine situation, the relationship between birth control and blood clots - or even COVID-19 and blood clots - hasn't been front page news.

As I was processing the back-and-forth about statistical risk and who was responsible for sharing what with whom, and at what level of amplitude, I couldn't help but think about all of the scholarship into the politics of numbers. We're living at a time when politicians are simultaneously espousing the need for "evidence-based policymaking" and working to diligently undermine, contort, or weaponize evidence. This is what scholars of "agnotology" mean when they talk about the manufacturing of ignorance through the seeding of doubt. Or what other scholars highlight as the "weaponization of transparency." 

I couldn't help but feel empathy for the scientists at J&J and the FDA who have been working around the clock trying to make a vaccine available to the public, trying to be responsible stewards of information and statistical risk in a context where their desire for caution can be turned on its head to undermine the legitimacy of their work. I also found myself feeling empathy for journalists who recognize the importance of reporting on this development, even as they know that their reporting is easily evolving into misinformation that's undermining the vaccine roll-out. Working with numbers is itself political.

To work in the world of medicine and science, statistics and probabilities is to grapple with trade-offs at a macro level, which present ethical conundrums even in the best of times. After all, that one terrible death from a blood clot could perhaps have been prevented by not taking the vaccine. But this is where we enter into the world of trade-offs, of unknowns, of morality. Without a vaccine rollout, many more people will die of blood clots from COVID-19. Had that woman been infected with COVID-19, she might have still succumbed to a blood clot. Medicine alters the dimensionality of risk. So how do ethics get negotiated? And by whom? This is the story of public health. 

Those complexities underpinning the advancement of science are complicated further by a politicized context such as that which surrounds the COVID-19 vaccine. Each act of communication can be twisted and contorted to convey different agendas, different values, different goals. Amplified transparency of risk is itself a political act. Sprinkle in the expectation in our current society that individuals are expected to make informed decisions for themselves, their families, and their communities, and we have a recipe for disaster. This is what the production of ignorance - aka misinformation, information disorder, agnotology, etc... - looks like in practice. The very acts of scientific transparency, which are intended to help inform decision-making, are twisted on their head, serving to undermining the legitimacy of scientific work and the coordination of a public that must work together to address a deadly disease. 

I keep wondering what it will take for the public to trust scientific information. But, perhaps, a better question might be: What kind of information is needed to help a fragmented public work together to solve societal-level challenges?

Note to the reader: These are questions that I'm struggling with. If you have thoughts, ideas (or even reading recommendations!), don't hesitate to reach out: zephoria [at] zephoria [dot] org.

danah