Twenty Times the CDC Exaggerated the Threat From Covid With False Data

The U.S. Centers for Disease Control and Prevention (CDC) published inaccurate data about the COVID-19 pandemic and made incorrect claims that exaggerated the threat on at least 20 occasions since January 2021, a new paper has found.

The pre-print (not yet peer-reviwed) by Vinay Prasad, Tracy Beth Hoeg, Kelley Krohnert and Alyson Haslam documents 25 instances when the CDC reported statistical or numerical errors. Twenty (80%) of these instances exaggerated the severity of the COVID-19 situation, three instances (12%) simultaneously exaggerated and downplayed the severity of the situation and one error was neutral. One error exaggerated COVID-19 vaccine risks. The CDC was notified about the errors in 16 instances (64%), and later corrected the errors, at least partially, in 13 instances (52%).

The authors searched for the errors by reviewing CDC publications, press releases, interviews, meetings and Twitter accounts. They also catalogued mortality data from both the National Center for Health Statistics and the CDC Covid Data Tracker and compared reported results.

They concluded that “a basic prerequisite for making informed policy decisions is accurate and reliable statistics, even during times of uncertainty”. They note a need for greater diligence in data collection and reporting. They also recommend that the federal entity responsible for reporting health statistics “should be firewalled from the entity setting policy due to concerns of real or perceived systematic bias in errors” – in this instance, towards exaggerating risk.

Here are the 25 errors they found:

February 26th 2021:

MMWR stated that during the study period, the seven-day moving average of cases identified by PCR or antigen testing ranged from 152 to 577 cases.
Multiple errors. Reported case rates during the study period were described as a seven-day moving average of cases per 100,000 persons including PCR and antigen cases, but the paper actually reported the raw seven day moving average (without adjusting for population) and for PCR only (not including antigen tests).
From: MMWR
Risk: Exaggerated
Concerns: Children

July 26th 2021:

Delta Variant is as contagious as chicken pox.
Delta is not as contagious as varicella. The CDC overstated Delta R0 and understated chicken pox R0 (Delta estimate was overlaid directly on a New York Times graphic).
From: CDC slide deck
Risk: Exaggerated
Concerns: All

July 27th 2021:

4% of COVID-19 deaths are in children 0-17.
Actual number was 0.04% based on original CDC estimated data. When the estimated data were updated later, the percentages were not updated. The actual percentage based on the updated data was 0.07%.
From: COVID-19 website
Risk: Exaggerated
Concerns: Children

October 15th 2021:

“COVID-NET data for the week ending Sept. 25th show that rates of COVID-19-associated hospitalisations in children ages 5-11 years are the highest they’ve been.”
COVID-NET hospitalisations were already falling from Sept peak. Rate was 1.1 in week ending Sept. 11th and Sept. 25th. (Now week of Sept. 11th shows 1.2),
From: Twitter @CDCgov
Risk: Exaggerated
Concerns: Children

October 27th 2021:

“CDC Director Walensky said “there have been 745 deaths in children less than 18.”
As of 27/10/21, NCHS data showed 558 deaths with COVID-19. Final NCHS data shows 679 pediatric deaths with COVID-19 through Oct. 30th, 2021
From: White House Press Briefing
Risk: Exaggerated
Concerns: Children

November 8th 2021:

Among ages 0-17, CDC’s reported rate of symptomatic illness was [more] than the total infection rate (asymptomatic + symptomatic –
an impossible claim), and this error occurred among children (infection rate also fell only for children from May 21st to Sept 21st estimates).
Estimated infection rate was 35,490 per 100K, not 29,885 per 100K (symptomatic illness remained at 30,253 per 100K).
From: COVID-19 website
Risk: Neutral
Concerns: All

December 20th 2021:

Omicron makes up 73% of new infections in the U.S.
Error with Nowcast estimate, a week later they revised to 23% (outside the previous 95% CI).
From: Data Tracker
Risk: Exaggerated
Concerns: All

February 24th 2022:

COVID-19 hospitalisations had a sudden over-1.6-fold increase in Georgia per HHS/CDC data.
Very likely a dramatic multi-week increase was due to imputation error on behalf of the reporting state or municipality, yet this was not audited or detected.
From: Data Tracker
Risk: Exaggerated
Concerns: All

March 15th:

Paediatric deaths on the Data Tracker demographics page were overstated while adult deaths were understated.
On 15/3/22, CDC removed 416 paediatric deaths from Data Tracker from 1,755 to 1,339 (still overstated) and almost 72,000 adult deaths, blaming an algorithm for classifying deaths as COVID-19 related.
From: Data Tracker
Risk: Mixed
Concerns: Both

June 17th 2022:

COVID-19 is a top five cause of death in children of all age groups.
Pre-print had inaccurate data, and CDC chose the most extreme version of the flawed data. Specifically, for COVID-19 it used cumulative counts (which spanned more than two years), and death was attributed if it was one of any multiple cause of death, whereas for other causes of death, they used only a single year, and attributed it only if it was the single underlying cause of death).
From: ACIP Meeting
Risk: Exaggerated
Concerns: Children

June 23rd 2022:

At a White House COVID-19 briefing, CDC Director Walensky cited the claim that COVID-19 is a “top five cause of death” in children
Flawed pre-print, authors already acknowledged that fact, and COVID-19 was not a top five cause of death.
From: White House Press Briefing
Risk: Exaggerated
Concerns: Children

June 27th 2022:

ACIP website includes the “top five cause of death” claim
Flawed pre-print, authors already acknowledged that fact, and COVID-19 was not a top five cause of death.
From: ACIP website
Risk: Exaggerated
Concerns: Children

August 9th 2022:

COVID-19 has killed 1,500 children ages 17 and younger.
As of 10/8/22, NCHS data showed 1,201 deaths with COVID-19. As of 5/2/23, NCHS data shows 1,323 paediatric deaths with COVID-19 through August 6th 2022.
From: Twitter @CDCgov
Risk: Exaggerated
Concerns: Children

August 12th 2022:

“COVID-19 hospitalisations for children and teens are increasing again in the U.S.”
CDC hospitalisation data showed hospitalisations had peaked two weeks prior, on 29/7/22.
From: Twitter @CDCgov
Risk: Exaggerated
Concerns: Children

August 20th 2022:

CDC Excess Mortality Dashboard overstated recent deaths in North Carolina and Connecticut.
Model for weighting due to death reporting lag was poorly adjusted.
From: CDC Excess Mortality Dashboard
Risk: Exaggerated risk of all-cause mortality
Concerns: All

August 22nd 2022:

Alabama paediatric hospitalisations had a dramatic single week increase from under 10 per day to over 50 per day.
Very likely a dramatic single week increase was due to imputation error on behalf of the reporting state or municipality, yet this was not audited or detected.
From: Data Tracker
Risk: Exaggerated
Concerns: Children

August 26th 2022:

CDC Data Tracker made a single week jump of 186 paediatric deaths and 1,679 adult deaths, which is unusually high for children and unusually low for adults.
Incorrect death data. CDC corrected this days later, removing 173 paediatric deaths and adding 2,484 adult deaths
From: Data Tracker
Risk: Mixed
Concerns: All

September 1st 2022:

ACIP Chair Grace Lee repeated the “top five cause of death” claim in ACIP meeting to approve bivalent booster.
Flawed pre-print was corrected two months prior. Unknown if ACIP committee informed.
From: ACIP meeting
Risk: Exaggerated
Concerns: Children

November 9th 2022:

Florida paediatric hospitalisations had a dramatic single week increase from seven to 112 (seven-day new admissions).
Very likely a dramatic single week increase was due to imputation error on behalf of the reporting state or municipality, yet this was not audited or detected.
From: Data Tracker
Risk: Exaggerated
Concerns: Children

December 30th 2022:

XBB.1.5 variant reported at 41% of new infections in the US.
A week later they revised to 18% (outside the original 95% CI).
From: COVID-19 website
Risk: Exaggerated
Concerns: All

December 31st 2022:

North Carolina paediatric hospitalisations had a dramatic single week increase from two to 19 (seven-day new admissions).
Very likely a dramatic single week increase was due to imputation error on behalf of the reporting state or municipality, yet this was not audited or detected
From: Data Tracker
Risk: Exaggerated
Concerns: Children

January 13th 2023:

Table 2 listed 62 events for children needing medical care as 13.9%.
It should be 1.9%. It is correct in the text, but not the table.
From: MMWR
Risk: Exaggerated risk of vaccine
Concerns: Children

February 9th 2023:

Dr. Walensky testified before Congress that there had been “2,000 paediatric deaths from COVID-19”.
This number comes from the flawed Data Tracker. Actual number is 1,400-1,500
From: Data Tracker/ testimony
Risk: Exaggerated
Concerns: Children

February 23rd 2023:

ACIP slide claimed 1,489 paediatric deaths in ages six months-17 years.
They did not remove 305 deaths in infants under-six months. Actual number should have been 1,184 using the NCHS data source cited on the slide
From: ACIP meeting
Risk: Exaggerated
Concerns: Children

Through March 3rd 2023:

Data Tracker continues to report too many paediatric deaths and too few adult deaths.
Inaccurate mortality data by age group are updated weekly on the CDC Data Tracker Demographics page.
From: Data Tracker
Risk: Mixed
Concerns: All

Read the full paper here.