Paused CDC program rattles maternal, infant health researchers

For nearly four decades, researchers have leaned heavily on a unique annual survey to understand the state of infant and maternal health in the United States. 

Called PRAMS, the dataset includes a richer variety of information about the life circumstances of newborns and their parents than birth certificates, providing insight into the causes of the higher rates of maternal mortality in the U.S. than in other developed countries. So researchers were distressed in recent weeks when they suddenly discovered they were locked out of the database, without any notice from the Centers for Disease Control and Prevention. And state health departments, which administer the survey, were told by the CDC to cease data collection. 

“If we can’t see this information, we cannot develop public health strategies to improve our maternal morbidity and mortality in the United States,” said Marian Jarlenski, a professor of health policy at the University of Pittsburgh. “I don’t know how else to say it, this data system is needed. It’s not an option. It’s part of having a functional public health system.”

The CDC told STAT on Friday that the interruption is temporary, and was the result of needing to make the survey compliant with some of President Trump’s executive orders. “PRAMS was not shut down,” Paul Prince, a spokesperson for the CDC, said in an email. “There were some schedule adjustments to ensure compliance with the Trump Administration’s Executive Orders, but these changes do not affect the continuation of the program.”

Prince declined to answer questions about what is being done to make the survey compliant or to specify the executive orders involved — although CDC has temporarily pulled down and made changes to other datasets to comply with Trump’s orders on DEI and gender. 

Last week, the agency posted a banner on the PRAMS website stating that 2023 data are “anticipated to be available to sites by March 2025,” and “PRAMS 2025 birth cohort data collection is anticipated to begin by April 2025.”

But, public health experts worry that the interruption could weaken an instrumental dataset, and it has made researchers reconsider their reliance on the survey over the long term. Researchers said that if the dataset or questionnaire were changed to delete mentions of race or ethnicity, it could harm efforts to improve maternal health. 

“If the race and ethnicity variables are removed, it would have major consequences for our ability to measure disparities in maternal and child health outcomes and to evaluate what works to ameliorate them. This would be very concerning given the glaring and persistent racial and ethnic disparities in many of the outcomes that PRAMS measures,” said Rebecca Gourevitch, a professor of health policy at the University of Maryland.  

The Pregnancy Risk Assessment Monitoring System survey is administered by state health departments before, during, and after pregnancy and is overseen by the CDC. The program is run in 46 states, the District of Columbia, Puerto Rico, and the Northern Mariana Islands, which account for up to 81% of births in the country. Researchers said the survey is uniquely powerful because of the range of topics it covers, from access to insurance and prenatal care to stressors and income. 

“We are, and have been, amid a maternal health crisis in the United States that has many different dimensions. The PRAMS dataset allows people to ask deep questions about those different dimensions,” said Katy Kozhimannil, a health policy researcher at the University of Minnesota. 

The dataset has been used to inform policy at a state and local level to guide the creation of laws to improve maternal health, and assess how successful they are. The data have also informed clinicians on how they practice and advise patients in terms of risk factors for adverse outcomes during a pregnancy, said Lindsay Admon, an OB-GYN and health policy researcher at the University of Michigan. 

For example, Admon and colleagues have used PRAMS data to show that many mothers whose births were covered by Medicaid end up uninsured 60 days after delivery, when their eligibility expired. Their work was used to support federal legislation to extend pregnancy-related Medicaid eligibility to a year postpartum, a change that has been adopted in 49 states and Washington D.C. 

The dataset occupies a niche of its own. Birth certificates lack granular data about the circumstances of the birth, and while insurance claims databases are more detailed, those can be costly to access. 

Researchers were startled by the interruption to the survey, in part, because it was done quietly. Jarlenski realized something was amiss when one of her students applied for access to the data, but did not receive a response from the CDC, not even a confirmation of the request. So, she tried to request the information, in case her student had done something wrong. but she also didn’t hear back. Meanwhile, she was able to get access to the CDC’s data on birth certificates. “That was really the red flag,” she said. 

One strength of the PRAMS dataset is that it has enabled policymakers to identify health inequities based on race, access to health care, or other factors in order to mitigate them. “There is nothing really biological that is contributing to the stark inequities and pregnancy outcomes and maternal outcomes — it has been produced by policies that are discriminatory,” said Joanna Almeida, a social epidemiologist at Simmons University who has pushed for the survey to include metrics of discrimination. “As such, the role of the government is to address them. [But,] if we don’t collect the data. We don’t know that the problem exists.”

Data are released on a delayed basis as they are validated, and researchers are eager to see the 2023 data, which would be the first to show any effects on maternal health from the Supreme Court’s 2022 Dobbs decision. 

Even if the CDC restores the database in the coming weeks, several researchers said their confidence in the program is somewhat shaken. But for a project of this scale, there are not many alternative groups that could support a replacement. And with that lack of information, there could be worse inequities in maternal health outcomes. 

Spurred on by the interruption, some have contemplated what alternatives to PRAMS might look like. While it was uncertain if the data would be made available again, the Data Rescue Project preserved topline data that was available on the CDC website. A survey on the scale of PRAMS would not be possible for any one organization to tackle, but some epidemiologists have begun discussions on how to create other, local datasets they can lean on. 

Jonathan Huang, an epidemiologist at the University of Hawaiʻi at Mānoa, has begun to have some conversations “to make sure that whatever happens, more people are made aware, can access, and can interact with these data.” Part of his goal, he said, is making people aware that the data exist to ask questions they may have, as well as support efforts that can answer more local questions.