If you are a patient looking to join a clinical trial, a clinician advising one, or a researcher trying to understand the trials that have been conducted for a particular medication, you may have encountered ClinicalTrials.gov. The website serves as a resource for clinicians and patients interested in trial participation as well as for researchers who want to understand the body of evidence for different conditions, assess signs of publication bias, and explore trial composition and representativeness. Since late 2007, federal law has required companies and researchers to register most Phase 2 to Phase 4 drug trials. Over 600,000 studies are listed on the site, which draws more than 3 million visitors each month and can be easily filtered based on patient age, sex, diagnosis, and ZIP code.
However, pregnant patients can’t easily identify studies for which they are eligible. ClinicalTrials.gov lacks structured data on whether pregnant or lactating people were included or can join a trial.
We believe the website should add a field to the participation eligibility criteria module that flags whether a study enrolls pregnant and/or lactating people and, if eligible, how many enrolled. This straightforward step would make trials easier to find and enable simple, accurate tracking of studies that enroll pregnant and lactating people.
Over the past decade, we have seen increasing concern about the lack of high quality medical evidence in pregnancy, particularly around drug safety. The shift began with efforts to reframe pregnant people from a “vulnerable” group requiring protections similar to children to a “scientifically complex” one, and bioethics work arguing why we should protect pregnant individuals “through” rather than “from” research. The issue entered the public consciousness through writers like Emily Oster and controversies around the exclusion of pregnant people from Covid-19 vaccine trials.
More recently, Trump administration officials have criticized the lack of drug safety evidence in pregnancy, particularly around the use of antidepressants and acetaminophen. Even as their scientific conclusions have been disputed, it remains striking that a politically diverse group of Americans agrees that more pregnancy studies are needed.
There are signs of progress, most notably ICH E21, a draft guideline from the International Council for Harmonisation, which would shift drug development from a presumption of excluding pregnant participants to a recommendation that sponsors plan early during drug development to collect data in pregnant and lactating people. The Food and Drug Administration released the draft for public comment in July 2025 alongside regulators in Europe, Japan, and other ICH regions, and final adoption is expected around early 2028.
Even amid these reasons for optimism, progress is not a foregone conclusion. ICH guidelines allow for expanded inclusion but don’t require it. Past calls for inclusion have led to minimal change. In prior work, one of us found that fewer than 1% of drug development clinical trials enroll pregnant patients, a statistic that didn’t budge over 15 years of data.
And obtaining that figure was unnecessarily difficult. We parsed data in ClinicalTrials.gov from 44,160 randomized drug trials open to women ages 18 to 45. To do this, we used large language models to read through the unstandardized fields, checked a sample against records we coded by hand, and called some of the study coordinators when the text didn’t say. About 76% addressed “pregnancy” in their inclusion or exclusion criteria reported as open text fields in the ClinicalTrials.gov database (0.8% explicitly permitted pregnant participants; 75% excluded them, e.g., through contraceptive requirements). The remaining 24% left pregnancy unspecified, which typically implied exclusion when we examined trial documents or called coordinators (who were uniformly very kind).
If we could calculate these numbers using LLMs, why bother adding additional required fields? Even with current technology, the data processing is laborious and error prone. And busy clinicians and sick patients who access the website are unlikely to be looking for a technical project.
Several expert panels have also emphasized the need for better data. The National Academies of Sciences, Engineering, and Medicine’s 2024 report “Advancing Clinical Research with Pregnant and Lactating Populations,” to which one of us contributed, noted that they could not use ClinicalTrials.gov to measure the extent of pregnant and lactating trial participation. And both that committee and the Task Force on Research Specific to Pregnant Women and Lactating Women, established by Congress in the 21st Century Cures Act, recommended improving data sharing and infrastructure, with the latter calling for a user friendly website for listing registries.
Adding these eligibility data to ClinicalTrials.gov would make it easier to find, fill, and track trials including pregnant and lactating patients. Clinicians and patients could more easily search for trials open to them. Researchers could study how enrollment evolves over time. And advocates and regulators could understand the impact of ICH E21 or other shifts in guidelines and regulations.
Improving ClinicalTrials.gov isn’t a panacea for the gaps in pregnancy data. It won’t address ethical and institutional review board concerns, resolve liability fears, or fund new research. But it’s a straightforward and inexpensive improvement at a time when better evidence for pregnant and lactating patients has rare bipartisan support.
ClinicalTrials.gov should add the fields.
Alyssa Bilinski is the Peterson family assistant professor of health services, policy, and practice and of biostatistics at the Brown University School of Public Health. Haley Sullivan is a faculty fellow at the Brown University School of Public Health. Anne C.C. Lee is the Levinger family professor of pediatrics at the Warren Alpert Medical School of Brown University.