JAMA Journal Publishes COVID Vaccine Study Squelched by CDC Leader
— Bhattacharya attacked study's test-negative design, a staple of vaccine effectiveness research
by Terrence Rudd, Staff Writer, MedPage Today
MedPabeToday reports:
Key Takeaways:
A study on COVID vaccine effectiveness that was suppressed by the CDC's acting director, due to its test-negative design, has been published in a major journal.
Test-negative designs have been used for decades by the CDC for routine influenza vaccine monitoring.
The case-control study showed that the estimated vaccine effectiveness rate was 50% against COVID-associated emergency department and urgent care visits and 55% against COVID-associated hospitalizations among adults.
“ The COVID vaccine study that was suppressed by the acting CDC director earlier this year has now been published in a JAMA journal.
The case-control study showed that the estimated vaccine effectiveness (VE) rate during the 2025-2026 COVID season's first 4 months was 50% (95% CI 42-57) against COVID-associated emergency department (ED) and urgent care visits and 55% (95% CI 41-66) against COVID-associated hospitalizations among adults, reported Ruth Link-Gelles, PhD, MPH, of the CDC's National Center for Immunization and Respiratory Diseases, and colleagues.
The numbers were slightly lower among adults 65 and older, with estimated VE rates of 48% (95% CI 37-56) against ED/urgent care visits and 53% (95% CI 37-65) against hospitalizations.
"The findings in this study demonstrate the added benefit of 2025-2026 COVID-19 vaccination irrespective of protection conferred by previous COVID-19 vaccination or SARS-CoV-2 infection," Link-Gelles and colleagues wrote in JAMA Network Openopens in a new tab or window. "The pace and frequency with which new SARS-CoV-2 lineages have become predominant, as exemplified by the spread of the SARS-CoV-2 variant BA.3.2, underscores the value of ongoing monitoring of COVID-19 VE and genomic surveillance."
The CDC estimates that the 2024-2025 COVID season led to 390,000 to 550,000 COVID-associated hospitalizations, with the greatest rate among adults ages 65 and up. As of February, CDC estimates put the number of COVID-associated hospitalizations at 99,000 to 180,000. Three recent studiesopens in a new tab or window showed that COVID vaccines were associated with lower risks of major adverse cardiovascular events, emergent care visits, hospitalizations, and critical illnesses, particularly among older adults.
To Test-Negative, or Not to Test-Negative
This study almost didn't see the light of publicationopens in a new tab or window. Originally slated to appear in March in the CDC's Morbidity and Mortality Weekly Report (MMWR), the study was rejected by the journal when the CDC's acting director and director at NIH, Jay Bhattacharya, MD, PhD, questioned its test-negative design that compared vaccination status in patients who tested positive for COVID (the cases) and those who tested negative for COVID (the controls). Bhattacharya also argued opens in a new tab or window that "MMWR is not currently a peer-reviewed journal -- but we are working on changing that."
Link-Gelles and colleagues explained that "the test-negative design has been used extensively to evaluate effectiveness of respiratory virus vaccines and provides a convenient and efficient method to rapidly evaluate effectiveness of new vaccine products in real-world settings (e.g., through EHR [electronic health] records)."
In an accompanying editorialopens in a new tab or window, Natalie Dean, PhD, of the Emory Rollins School of Public Health in Atlanta, said that the test-negative design is an approach that works because patients who test negative for SARS-CoV-2 can approximate the source population in which COVID cases arise. That allows researchers to estimate VE without the burden of establishing a full population denominator.
"The TND [test-negative design] avoids the need to establish a fully enumerated cohort of vaccinated and unvaccinated populations followed in time," Dean wrote. In addition, "unlike case-control studies that sample healthy community controls, the TND can be implemented efficiently because it leverages controls who are already seeking care at the same sites."
The use of a test-negative design by the CDC is nothing new. It's been used for decades for routine influenza vaccine monitoring, Dean pointed out, and it "underpins seasonal influenza vaccine estimates published by the CDC."
In nations like the U.S., where large healthcare databases can't be leveraged to assess VE, "the TND offers a pragmatic solution to the tricky problem of estimating vaccine effectiveness," she added.
Researchers aren't ignoring the limitations of the test-negative design highlighted by critics like Bhattacharya, Dean countered. Those limitations include confounding by health-seeking behavior, in which people who are most likely to get vaccinated are also more likely to seek medical treatment when they're symptomatic. But because the design restricts its test-negative controls to those people who seek care, she said, the approach has an advantage over other observational study designs.
"It is standard for TND studies to be accompanied by extensive sensitivity analyses, assessing the potential for bias, testing for residual confounding, and checking the robustness of the interpreted findings," she wrote. "The continued refinement of these methods reflects the field's commitment to improving routine vaccine monitoring."
Because routine VE estimates are vital to public health decision-making, Dean noted, "the TND remains an important and practical approach while other options continue to mature."
The case-control study used the Virtual SARS-CoV-2, Influenza, and Other Respiratory Viruses Network (VISION) and included data from 85,725 ED/urgent care visits and 26,073 hospitalizations at 253 EDs/urgent care centers and 179 hospitals in seven states from September through December 2025. All patients had COVID-like illness and were tested for SARS-CoV-2 in the 10 days before or 72 hours after an ED/urgent care visit or hospital admission.”
My comment:
Bolded emphasis mine
by Terrence Rudd, Staff Writer, MedPage Today
MedPabeToday reports:
Key Takeaways:
A study on COVID vaccine effectiveness that was suppressed by the CDC's acting director, due to its test-negative design, has been published in a major journal.
Test-negative designs have been used for decades by the CDC for routine influenza vaccine monitoring.
The case-control study showed that the estimated vaccine effectiveness rate was 50% against COVID-associated emergency department and urgent care visits and 55% against COVID-associated hospitalizations among adults.
“ The COVID vaccine study that was suppressed by the acting CDC director earlier this year has now been published in a JAMA journal.
The case-control study showed that the estimated vaccine effectiveness (VE) rate during the 2025-2026 COVID season's first 4 months was 50% (95% CI 42-57) against COVID-associated emergency department (ED) and urgent care visits and 55% (95% CI 41-66) against COVID-associated hospitalizations among adults, reported Ruth Link-Gelles, PhD, MPH, of the CDC's National Center for Immunization and Respiratory Diseases, and colleagues.
The numbers were slightly lower among adults 65 and older, with estimated VE rates of 48% (95% CI 37-56) against ED/urgent care visits and 53% (95% CI 37-65) against hospitalizations.
"The findings in this study demonstrate the added benefit of 2025-2026 COVID-19 vaccination irrespective of protection conferred by previous COVID-19 vaccination or SARS-CoV-2 infection," Link-Gelles and colleagues wrote in JAMA Network Openopens in a new tab or window. "The pace and frequency with which new SARS-CoV-2 lineages have become predominant, as exemplified by the spread of the SARS-CoV-2 variant BA.3.2, underscores the value of ongoing monitoring of COVID-19 VE and genomic surveillance."
The CDC estimates that the 2024-2025 COVID season led to 390,000 to 550,000 COVID-associated hospitalizations, with the greatest rate among adults ages 65 and up. As of February, CDC estimates put the number of COVID-associated hospitalizations at 99,000 to 180,000. Three recent studiesopens in a new tab or window showed that COVID vaccines were associated with lower risks of major adverse cardiovascular events, emergent care visits, hospitalizations, and critical illnesses, particularly among older adults.
To Test-Negative, or Not to Test-Negative
This study almost didn't see the light of publicationopens in a new tab or window. Originally slated to appear in March in the CDC's Morbidity and Mortality Weekly Report (MMWR), the study was rejected by the journal when the CDC's acting director and director at NIH, Jay Bhattacharya, MD, PhD, questioned its test-negative design that compared vaccination status in patients who tested positive for COVID (the cases) and those who tested negative for COVID (the controls). Bhattacharya also argued opens in a new tab or window that "MMWR is not currently a peer-reviewed journal -- but we are working on changing that."
Link-Gelles and colleagues explained that "the test-negative design has been used extensively to evaluate effectiveness of respiratory virus vaccines and provides a convenient and efficient method to rapidly evaluate effectiveness of new vaccine products in real-world settings (e.g., through EHR [electronic health] records)."
In an accompanying editorialopens in a new tab or window, Natalie Dean, PhD, of the Emory Rollins School of Public Health in Atlanta, said that the test-negative design is an approach that works because patients who test negative for SARS-CoV-2 can approximate the source population in which COVID cases arise. That allows researchers to estimate VE without the burden of establishing a full population denominator.
"The TND [test-negative design] avoids the need to establish a fully enumerated cohort of vaccinated and unvaccinated populations followed in time," Dean wrote. In addition, "unlike case-control studies that sample healthy community controls, the TND can be implemented efficiently because it leverages controls who are already seeking care at the same sites."
The use of a test-negative design by the CDC is nothing new. It's been used for decades for routine influenza vaccine monitoring, Dean pointed out, and it "underpins seasonal influenza vaccine estimates published by the CDC."
In nations like the U.S., where large healthcare databases can't be leveraged to assess VE, "the TND offers a pragmatic solution to the tricky problem of estimating vaccine effectiveness," she added.
Researchers aren't ignoring the limitations of the test-negative design highlighted by critics like Bhattacharya, Dean countered. Those limitations include confounding by health-seeking behavior, in which people who are most likely to get vaccinated are also more likely to seek medical treatment when they're symptomatic. But because the design restricts its test-negative controls to those people who seek care, she said, the approach has an advantage over other observational study designs.
"It is standard for TND studies to be accompanied by extensive sensitivity analyses, assessing the potential for bias, testing for residual confounding, and checking the robustness of the interpreted findings," she wrote. "The continued refinement of these methods reflects the field's commitment to improving routine vaccine monitoring."
Because routine VE estimates are vital to public health decision-making, Dean noted, "the TND remains an important and practical approach while other options continue to mature."
The case-control study used the Virtual SARS-CoV-2, Influenza, and Other Respiratory Viruses Network (VISION) and included data from 85,725 ED/urgent care visits and 26,073 hospitalizations at 253 EDs/urgent care centers and 179 hospitals in seven states from September through December 2025. All patients had COVID-like illness and were tested for SARS-CoV-2 in the 10 days before or 72 hours after an ED/urgent care visit or hospital admission.”
My comment:
Bolded emphasis mine
