In what is perhaps the pinnacle of health policy self-sabotage so far in this pandemic, the CDC announced yesterday updated recommendations for a return to masking, motivated by fears of slight surges in COVID-19 cases and the Delta variant. CDC Director Rochelle Walensky reiterated that while most COVID transmission is between unvaccinated individuals,

“information on the Delta variant from several states and other countries indicate that in rare occasions, some vaccinated people infected with the Delta variant after vaccination may be contagious and spread the virus to others. This new science is worrisome and unfortunately warrants an update to our recommendations.”

This would be some concerning new data indeed, except Walensky never cited what data – from “several” states and countries – she was referring to, and of course, not one question in the briefing asked her to clarify. No updates to the CDC website including this data or any follow-up press releases to name the studies in particular have been released.

There is some speculation by AP, and others, that the concern of the CDC is based on one Chinese study made public earlier this month. The study, currently unpublished and not peer-reviewed, looks at 126 samples and assumes transmissibility by finding a 1260 times higher viral load. The study seems to not account for or include information on the vaccination status of the individuals, which vaccine they might have had, nor does it empirically connect the transmissibility or infectiousness of Delta to the increased viral load – it is inferred.

We do know that there are currently multiple peer-reviewed studies that do specifically trace the transmissibility of COVID-19 in vaccinated individuals. One study, released two days before the Chinese report, follows over 100,000 individuals with the Moderna vaccines, and calculates that the vaccine reduces transmissibility by “at least 61%.” Again, this is transmissibility – the probability the virus will be passed on to you – not vaccine efficacy, which would deal with the percentage reduction in infected (which we know to be over 90% for Moderna and Pfizer vaccines).  Another article in pre-print, though it should be noted it is an update of a previous paper, examines and collates the literature on the transmissibility of COVID-19 among vaccinated individuals. The meta-analysis of 33 studies shows, among others, that

“Evidence from two large household surveillance studies from the UK suggests that a single or full dose of AstraZeneca (AZ) and Pfizer-BioNtech (PfBnT) vaccines may prevent household transmission of COVID-19 after 14 days of vaccination by up to 54%.”

Many of the other included studies report prevention of transmission from 59% to 90%, depending on exposure as well as symptomatic vs asymptomatic infections. Seven included studies discovered lower viral loads in vaccinated individuals when compared to unvaccinated individuals.

So, let’s go back to the CDC and Director Walensky. One would hope that a grand statement that flies in the face of all previous evidence regarding transmission, viral loads, infection rates, and vaccines would be backed by significant data and peer-reviewed publications.

Nope.

Dr. Walensky never shows her work, cites no sources, and leaves it up to the journalists to investigate her claims. Obviously, few do, and even then the AP can only find one small study without peer-review that they assume these guidance changes could be based on. Taking an authority figure by their word, in the face of overwhelming evidence to the contrary with no stated evidence to the contrary, is not science. It is faith.

This is not the first time the CDC Director has taken liberties in making broad moralizing statements without providing any evidence, and it is not nearly the first time the press has simply failed to ask for that data.

In the White House briefing on July 16th, Walensky infamously stated that we have a “pandemic of the unvaccinated” and cited that nearly 97% of hospitalizations for COVID are unvaccinated individuals. This followed on the heels of another – unchallenged – statement by Dr. Fauci on “Meet The Press” that 99% of hospitalizations are caused by the unvaccinated. Neither individual backs up how they arrived at these figures, what their methodology was, or what data they might actually be citing (are we dealing with recent cases? All cases since vaccinations started? Are we including all COVID positives regardless of the reason for being hospitalized?).

The press, of course, ran with the rhetoric and attempted to create or devise their own methodologies for coming to similar conclusions. The AP states in one article they did their calculations in-house and then proceeds to show basic division and subtraction to conclude their figures. Unfortunately, they and many other outlets calculated a percentage of vaccinated vs unvaccinated, but it is an essentially useless number. It is presented by these outlets as a measurement of risk, except it can’t possibly be – as it ignores accounting for baseline risks. Journalists are not statisticians and should not pretend to be. The Washington Post gives a little more information for their own data visualizations, describing their methodology as follows:

But this methodology includes information from the CDC, apparently related to the 97% statement, without stating what – or where – that information is. There is no link, no data set, and no citations to the CDC data – which seems to have never publicly existed. Even in the Powerpoint notes for Walensky’s July 16 briefing, there are no citations that might suggest where the data they use is coming from, or what methodology was used to calculate that data.

Again, Walensky wants us to take her word on faith alone, and many do.

The thing is, Walensky maybe be right. Evidence may well exist or emerge that shows the Delta variant is just as transmissible in vaccinated individuals or that 97% of current COVID-19 hospitalizations are from the unvaccinated. But without the data and methodology, there is nothing to back her claims, and zero ability to critique those claims. For instance, does the 97% figure control for socioeconomic determinants of health, or does it assume each individual has equal access to healthcare? Does it account for demographic variation and compare the ill individuals to healthy individuals within those demographic constraints? Does the number account for already terminal patients who find the vaccine unnecessary?

These are just a few of the very valid and expected considerations when one wants to crunch this kind of data, yet all we are left to do is assume. We’re expected to have a blind trust, blind faith, in the same people who refuse to cite their own sources. Based on these figures, Walensky unhelpfully suggested that “nearly every death, especially among adults, due to COVID-19, is, at this point, entirely preventable.” Behind Walensky’s sleight of hand, there is a suggestion that the percentage of unvaccinated in hospitals extrapolates to the extent to which vaccination reduces risk. That, however, is entirely scientifically unsound. When comparing risks, finding a suitable comparator is crucial. This ignores baseline risks – persons who are unvaccinated for various reasons, often outside their control, might also be predisposed towards worse clinical outcomes. America deserves truth and science from the CDC, not inappropriate risk measurements and semantic trickery.

The CDC Director’s unwillingness to be transparent, up-front, and open about her data and methods is not only concerning on a critical scientific level – it has also resulted in one disastrous implosion of public health policy and communications after another. The head of the CDC just threw out random numbers that have not been published in anything of repute, including the CDC’s own peer-reviewed journal (the Mortality and Morbidity Weekly Reports), and expect people to not only take it seriously – but adjust their lives based on it. These same people and institutions who nag about the lack of public trust in them by the public then go on to become such enormous sources of misinformation it would give Facebook and Twitter whiplash.

I remember the days when Trump took a lot of mockery and criticism for drawing over a hurricane’s paths with a sharpie.

How is this any better?