I've lost all faith in the CDC's leadership

Prevention and Control is in the name, and yet

May 05, 2022

The CDC held a media briefing session on April 26, in which Director Walensky gave an update on the status of the pandemic. I’ve responded to some of her remarks below. If you don’t have time for the whole post, please jump down to the section titled, “Proving a reliable picture?”

Overall, we continue to have some mixed trends as seen on slide 1, deaths fortunately are continuing to trend downward with a 70 average of about 300 per day, which represents an, um, estimated 18% decline from the prior week.

Here’s the current state.

Hospital admissions also remain low at about 1600 per day, but we should note that for the second week in a row, they are slowly trending upwards and increase of about 9%. this week from the prior week.

Here’s where that stands.

Cases also remain comparatively low to where we were even a month ago at about 44,000 per day. Though, this too has represented an increase of about 25%

The 7-day rolling average hit 61,513 on May 4. (The daily count was over 115k.)

We are aware that many are using rapid antigen tests that may not be reflected in the total daily case counts. Our data show that we are still doing over 650,000 PCR tests daily. And while this is about a fivefold reduction in PCR testing then was done during the omicron peak in January and February, we are now recording 70 fold fewer cases suggesting that our reported cases counts do reflect a true and reliable drop in our overall cases.

Since peaking during the Omicron surge in January, the number of tests has steadily fallen.

Meanwhile, test positivity has climbed since March 17. A significant increase in test positivity is evidence of a surge. We’re giving far fewer recorded tests (1.7/1000 people now vs. 7.8/1000 people at Omicron’s peak), so we may be fostering a situation where we’re getting bias in the data.

If testing is going down, and what remains is tilting towards the affluent, the data may be hiding a worse overall picture due to the sampling. I’ll caution this line of thinking by noting that I’m not a scientist, but we know that “people in poorer (US) counties have died overall at almost twice the rate of those in richer counties.”

Proving a reliable picture?

We’re of course, continuing to follow trends in positive PCR tests and have also looked at data on a selected share of rapid tests. The trends are largely similar, and we continue to believe that our PCR testing data, especially when we corroborate it with information from our other surveillance systems like wastewater surveillance and emergency department and hospital surveillance provide us a reliable picture of the trajectory of COVID-19 across our country.

I learned about the wastewater data tracking efforts from a researcher I met at a conference late last year. He explained how it works and that it’s a leading indicator (as opposed to the lagging one we get from hospitals) that can help governments respond quicker to outbreaks. I was happy to learn that the CDC had added that to its COVD dashboard, but I’ve struggled with something.

Here’s the latest Community Levels map from the CDC. It’s the new ‘Don’t worry, be happy’ COVID-19 tracker that tells 90% of Americans that all is well.

And here’s the older Community Transmission map that was previously used to keep all of us informed, but now is just recommended for use by hospital facilities. You may notice a slight difference between the two, but you’ll see that New York state is not doing well on both maps.

Edit: I just ran across this tweet that calls out the shocking difference between the two maps.

Let’s revisit the last comment from Director Walensky.

The trends are largely similar, and we continue to believe that our PCR testing data, especially when we corroborate it with information from our other surveillance systems like wastewater surveillance and emergency department and hospital surveillance provide us a reliable picture of the trajectory of COVID-19 across our country.

Here’s the wastewater surveillance data from the CDC. Sites with no recent data are excluded to avoid muddying the picture.

And here’s a close-up of New York.

This map provides a measure of the percent of tests that included any trace of COVID-19.

For comparison, here’s the state of New York from the Community Transmission map above. All but a handful of New York’s test sites show 0% of their tests having the presence of COVID-19, while all but a handful of the state’s counties are rated High in the Community Transmission map (and the rest are rated Substantial.)

How is this possible? Are the people with COVID-19 in most New York counties on septic tanks or using outhouses? Or does the CDC have a mess in the data that Director Walensky claims it is using to “provide us a reliable picture of the trajectory of COVID-19 across our country”?

Note: This post was initially titled I’ve lost all faith in the CDC, which was not an accurate reflection of my position. I’m confident there are loads of scientists and people in other roles working hard to prevent harm, but the organization’s leadership has lost me.