The C.D.C. Needs to Stop Confusing the Public

By Zeynep Tufekci in The NY Times. Dr. Tufekci is a contributing Opinion writer who has extensively examined the Covid-19 pandemic. Excerpts:

"In announcing changes in mask recommendations Dr. Walensky notably said that vaccinated people who became infected had viral loads similar to those of unvaccinated people who got sick, and could “forward transmit with the same capacity as an unvaccinated person.”

That vaccinated people with breakthrough infections could sometimes transmit the virus wasn’t particularly surprising given the data and anecdotes that had accumulated. However, that they had “the same capacity” to transmit it as an unvaccinated person certainly was, including to many experts. It was the kind of claim that was clearly going to alarm tens of millions of vaccinated people and needed to be delivered with maximum clarity and context.

Instead we got a stark lesson in how not to communicate.

First, the data that the C.D.C. said it based its decision on wasn’t released right away — leaving both experts and ordinary people to try to piece together what was being said.

The Associated Press, on the same day as Dr. Walensky’s news conference, quoted her as saying the level of virus in infected vaccinated people was “indistinguishable” from the level of virus in the noses and throats of unvaccinated people. The news report also noted that the data the C.D.C. had based its new decision on was unpublished, and had “emerged over the last couple of days from over 100 samples from several states and one other country.”

Which ones and where?

The updated C.D.C. guidelines pointed to a single reference on this question, which was a preprint looking at health care workers in India who became infected after receiving vaccines not approved for use in the United States. In many studies, viral load is ascertained through examining a measure from PCR testing called the cycle threshold, or CT. While useful, because the cycle threshold is a proxy and a snapshot, it needs to be interpreted with caution. Still, the study from India wasn’t even between those who were vaccinated and those who were not, but among vaccinated workers infected with different variants. That the more transmissible Delta had a higher viral load than other variants had already been reported.

Two days later, based on slides leaked from the C.D.C. and a federal health official who spoke on condition of anonymity, The Washington Post reported that the American outbreak that the agency used to base its new guidelines on was in Provincetown, Mass.

Finally, on July 30, the C.D.C. released its epidemiological study of the Provincetown outbreak.

In a seashore town with about 3,000 residents, the vast majority of whom were vaccinated, and 60,000 summer visitors, the C.D.C. noted more than 450 infections between July 3 and 17. Health officials later traced more than 960 cases to gatherings in Provincetown. Attendees said it rained a lot during those two weeks, driving more people to crowded, poorly ventilated bars and restaurants, probably worsening the spread.

The data showed what had been documented elsewhere: Delta was sometimes able to infect the vaccinated, although there were only seven hospitalizations and no deaths. The most common symptoms were cough, headache and sore throat.

The cycle threshold numbers among vaccinated and unvaccinated cases were indeed similar, suggesting similar viral loads — at least when the test was taken. However, since there was no follow-up to measure actual transmission events, it was unclear how much the vaccinated contributed to the spread. Or did the unvaccinated infect most people in the outbreak, including the vaccinated? It is also hard to draw broader conclusions from such nonsystematic and limited data. People who get tested are likely to be sicker, for example, so they probably have higher viral loads to begin with, and thus are probably not fully representative of those who are vaccinated but infected. All this means that the cycle threshold value may be useful, but it’s just one piece of the puzzle without contextual data — suggestive but not conclusive.

On July 31, a more systematic study from Singapore showed that viral load from Delta could get high but that it quickly peaked and then crashed in breakthrough cases among the vaccinated — as their immune system responded to quickly clear the virus. The potential infectious period lingered much longer in the unvaccinated."

"the study should not have been presented as the primary cause for the alarm it set off, and the public certainly should not have been left waiting many days for the data itself while details leaked out in dribs and drabs, often through anonymous sources."