Medication in pregnancy: Better data for healthier kids

Researchers need much bigger data sets to find outcomes that matter, Dr. Sascha Dublin and colleagues write in Pediatrics.

Researchers at Kaiser Permanente Washington Health Research Institute (KPWHRI) recently called for substantially larger data sets and higher quality data for studies of prescription drug use in pregnancy to improve children’s health. 

“Given that the vast majority of pregnant women take at least one medication, we know concerningly little about the impact of these exposures on children, both short- and long-term,” wrote Senior Investigator Sascha Dublin, MD, PhD, and colleagues in a commentary in the June 8 Pediatrics, “Studying Medication Safety in Pregnancy: A Call for New Approaches, Resources, and Collaborations.”

The team pointed to a large study of antidepressants in pregnancy led by Gretchen Bandoli, PhD, at the University of California, San Diego (UCSD) that appeared in the same issue. Despite the researchers’ access to a data set of about 200 million lives, their source included just 15,000 pregnancies exposed to antidepressants and only 434 with sustained high-dose exposure—"too few to draw meaningful conclusions,” wrote Dr. Dublin. 

How can such problems be solved?  With “new structures and policies that incentivize collaboration and data sharing,” Dr. Dublin wrote. “We need bold visions, funding mechanisms, and governance policies to stimulate and support large transnational consortia that can generate data sets of 10 to 20 million pregnancies.”

Changes are also needed to improve data quality. Insurance claims data from medical visits can be used to achieve large samples, but they lack characteristics that can be found in other sources, such as electronic medical records and interviews—information such as weight, smoking status, or use of alcohol or over-the-counter drugs, Dr. Dublin explained.

Mobile apps and Internet-based questionnaires hold promise for efficient data collection in large populations, Dublin suggested. She points to her research group’s collaboration with the U.S. Food and Drug Administration on a mobile app that was pilot tested in pregnant women. Conducting large interview-based studies within health care systems with strong track records of using their electronic health records for research is another approach that could be fruitful, she and her colleagues wrote.

Dr. Dublin’s co-authors were KPWHRI Research Associate Paige Wartko, PhD, MPH; and Rita Mangione-Smith, MD, MPH, executive director and senior investigator at KPWHRI and vice president of research and health care innovation at Kaiser Permanente Washington.

“Because in utero and early childhood experiences provide the foundation for long-term health, it is imperative that we generate better information about the impact of medication exposures on infant and child outcomes,” they concluded.

--by Joan DeClaire