An analysis leveraging artificial intelligence (AI) has revealed a concerning lack of representation of pregnant individuals in clinical trials. The study, which assessed over 40,000 clinical trials, found that less than 1% included pregnant participants, a rate that has remained virtually unchanged over the past 15 years despite increasing calls for more inclusive research practices. This gap in data leaves pregnant individuals and their healthcare providers with limited evidence-based information for making critical medical decisions.
The researchers utilized AI to analyze the inclusion and exclusion criteria of a vast number of clinical trial records. The AI model was trained to identify specific language indicating the inclusion or exclusion of pregnant individuals, as well as related criteria such as negative pregnancy tests or postmenopausal status. The AI was designed not only to classify the trials but also to provide supporting evidence for its classifications, enhancing the reliability and transparency of the results.
AI-Driven Methodology
The AI model was developed through a rigorous process of test-driven development. Researchers initially trained the AI on a smaller set of studies and refined its accuracy through multiple iterations. A second AI agent was used to review the results of the first, identifying and correcting errors. Ultimately, the model achieved an accuracy rate exceeding 98% after human review of 1,000 studies used as a larger training set.
According to the researchers, without AI, the scope of the study would have been significantly limited. "If we didn't have AI, we would have only been able to look at a much smaller sample of trials, and we would have had a much less well rounded understanding of this phenomenon," they stated. The AI-driven approach allowed the team to analyze the full sample in approximately one hour, a task that would have taken a research assistant an entire summer to complete manually.
Implications for Pregnant Individuals
The persistent exclusion of pregnant individuals from clinical trials raises significant ethical and practical concerns. Without adequate data on the safety and efficacy of medications during pregnancy, healthcare providers and pregnant individuals are forced to make decisions based on incomplete or extrapolated information. This can lead to suboptimal treatment choices and potential harm to both the pregnant individual and the developing fetus.
"We’re asking women to make a hard call during a really important time, with less data than people would normally have available to them to make decisions about taking medications," the researchers noted. They advocate for greater inclusion of pregnant individuals in clinical trials to address this critical evidence gap.
Historical Context and Future Directions
The exclusion of pregnant individuals from clinical trials has historical roots. Until 1962, the FDA did not require companies to submit evidence of medication safety and efficacy. For many years, women of childbearing age were often barred from participating in clinical trials. While federal law in 1993 mandated the inclusion of non-pregnant individuals and broader representation in clinical trials, the inclusion of pregnant individuals has lagged behind.
The researchers express hope that in the future, the exclusion of pregnant individuals from clinical trials will be viewed as anachronistic, similar to the historical exclusion of women. They emphasize the importance of cautious but engaged use of AI in research, advocating for test-driven development to ensure the reliability and accuracy of AI-generated results.
