From maternal health to influenza rates, gaps in CDC’s public health data are creating dangerous blind spots for disease tracking and prevention

Public health relies on data – whether it is tracking the effectiveness of a given year’s flu vaccine, monitoring blood lead levels around the country or estimating the prevalence of diabetes. This data forms the basis for decisions such as whether a community should expand screening for diabetes and which communities are at greatest risk of severe flu-related illnesses.

But in January 2025, at the beginning of President Donald Trump’s second term, webpages and data from the Centers for Disease Control and Prevention – and from many other federal agencies – began to disappear. In total, at least 200 CDC datasets and more than 8,000 webpages from across the government were taken down.

Though many of these resources were subsequently restored in the months that followed, the speed and scale of the removals make a full accounting difficult. Even now, more than a year after so much data from federal agencies was removed, researchers like me are discovering new mechanisms through which this same data remains at risk.

Most recently, for example, more than 6,000 webpages on energy conservation wwere removed from federal websites while a heat wave ravaged the U.S. through Fourth of July celebrations.

I am an infectious disease epidemiologist who studies how respiratory pathogens affect communities around the world and how data is used to monitors these impacts. Every day, health departments, practitioners and researchers have to make decisions on where to direct resources, which conditions to monitor and even which research questions to prioritize. Without high-quality data to support them, these decisions run the risk of becoming mere guesses.

Interruptions in data collection

Collecting data, in partnership with state, tribal and local public health experts, is among the most important functions of the CDC. In fact, nearly 80% of the the agency’s funding goes toward these collaborations to produce critical data that the entire public health system rests on.

A January 2026 study highlighted a troubling trend that began in early 2025. Among CDC data that had previously been updated monthly, nearly half – 46% – had unexplained pauses in such updates.

The vast majority of these pauses – about 90% – are related to data on vaccinations or impacts of common respiratory viruses, such as influenza, COVID-19 and respiratory syncytial virus, or RSV. Considering the outsize roll these pathogens play in driving severe respiratory illnesses, such as pneumonia, these pauses are alarming.

It’s not just respiratory disease data that is affected. The CDC also suspended updates for data on HIV incidence and prevalence, as well as on whether a person knows they’re infected. The agency’s website previously cited staffing reductions as the cause.

It’s unclear which of these datasets have resumed collection. Some data, such as estimates of HIV incidence and prevalence, had still not been updated as of July 6, 2026, though a banner on the webpage states that “CDC is updating these data.” And other data, such as that on emergency department visits for respiratory diseases, appears to have been updated again.

However, the lack of clear statements from the CDC on the status of different data sources makes it difficult to monitor where they stand.

Data loss via staff cuts

Widespread staff cuts starting in early 2025 across the Department of Health and Human Services, which houses the CDC, crippled much of the agency’s data collection and dissemination.

Some of the agency’s data is publicly accessible. But to access some particularly sensitive CDC datasets, researchers must also apply and go through a vetting process. If federal workers who do that vetting are terminated, the data effectively becomes inaccessible.

That’s what happened in the case of the Pregnancy Risk Assessment and Monitoring System, the gold standard data source on maternal and child health in the U.S. In January 2025, the CDC quietly stopped processing data access requests. Then, months later, on April 1, the team that oversaw this program was terminated.

Pregnancy Risk Assessment and Monitoring System data from before 2016 remains available, but the only way researchers can access data from 2016 onward is to submit individual requests to the 46 participating states, Washington, D.C., Puerto Rico or the Northern Mariana Islands. This is an enormous burden for both researchers looking to work with this data and for the state, local and territorial health departments, which may be ill-equipped to process these requests.

Later in 2025 the administration conducted more targeted terminations in offices tasked with planning future data collections. In October 2025 two critical teams were fired: one in charge of planning the National Health and Nutrition Examination Survey, which has collected key health and nutrition metrics continuously since 1999, and the other in charge of managing the National Death Index, which compiles death record information.

The nutrition database, much of which is publicly accessible, informs policies and decision-making related to food labeling, dietary guidelines and many other topics. The death index, access to which is highly regulated, is a crucial resource for health and medical researchers for studying causes of death.

Though staff terminations relating to both these CDC datasets were reversed within a few months, the prospect of losing access to already existing and yet-to-be-acquired data in these collections sent shock waves of concern through the public health research community.

Loss of trust

For the CDC to do its work of monitoring and protecting public health, it must build and maintain a great deal of trust, both with the people and communities providing personal health data and with the state and local health agencies often responsible for collecting it.

In my view, HHS’s removal and alteration of CDC data damages that trust. Concerns about the agency’s inappropriate access to and use of sensitive data are continuing to emerge. Health departments – as well as individuals – may hesitate to provide data to the CDC, or even seek essential health services, if they fear it will be shared in inappropriate ways without their consent.

Such data irregularities fit into a broader pattern of disrupting long-standing agency activities – for example, abruptly terminating grants to state and local health departments and disrupting health messaging on vaccines.

While the CDC and its staff continue to partner with state and local agencies to support public health in the U.S., these trends threaten to destroy hard-won trust that would be difficult to rebuild.