Clinical Report: Utilizing Machine Learning to Address Evidence Deficiencies in Medication Safety for Pregnant Women

Overview

This report discusses the use of machine learning to bridge evidence gaps in medication safety for pregnant women, who have historically been excluded from clinical trials. It highlights the importance of interpretability in machine learning models to ensure reliable conclusions.

Background

The exclusion of pregnant women from clinical research has led to significant evidence gaps regarding medication safety, impacting the 79% of women who take medications during pregnancy. Despite regulatory efforts to include women in research, the trend of exclusion persists, necessitating innovative approaches to assess drug safety. Machine learning offers a promising avenue to analyze large datasets and improve understanding of medication risks for this population.

Data Highlights

No specific numerical data or trial data was provided in the article.

Key Findings

• 79% of pregnant women take at least one medication, often without sufficient evidence on safety.
• Only 4% of clinical trials in the past decade included pregnant women.
• Machine learning can analyze large datasets to identify links between medication exposure and outcomes.
• Transparent machine learning models can help establish causality and improve drug safety recommendations.
• Inclusion of pregnant women in clinical trials could have prevented significant adverse outcomes, such as thalidomide-related birth defects.

Clinical Implications

Healthcare providers should be aware of the evidence gaps in medication safety for pregnant women and consider the insights from machine learning analyses when making treatment decisions. The integration of causal inference methods with machine learning can enhance the reliability of findings related to drug safety in this vulnerable population.

Conclusion

Addressing the evidence deficiencies in medication safety for pregnant women is crucial for improving clinical outcomes. Utilizing machine learning with a focus on transparency and causality can significantly enhance our understanding of medication risks during pregnancy.

Related Resources & Content

1. Drug Safety, 2024 -- Developing a New Algorithm for Detecting Adverse Drug Reactions During Pregnancy in Pharmacovigilance Systems: Insights from EudraVigilance Data
2. npj Digital Medicine, 2025 -- Machine Learning and Network Toxicology Approaches for Computational Pharmacovigilance of Lifitegrast in Managing Dry Eye Disease
3. Drug Safety, 2022 -- Utilizing Machine Learning for Causal Analysis in Pharmacovigilance Applications
4. Drug Safety, 2018 -- Navigating Safety Data: Employing Machine Learning to Detect Individual Case Safety Reports in Social Media Platforms
5. ICH E21 guideline on inclusion of pregnant and breastfeeding individuals in clinical trials – Scientific guideline | European Medicines Agency (EMA), 2025
6. Acetaminophen Use in Pregnancy and Neurodevelopmental Outcomes | ACOG, 2025
7. AI Drug Safety in Pregnancy | Pregnancy | JAMA | JAMA Network, 2025
8. ICH E21 guideline on inclusion of pregnant and breastfeeding individuals in clinical trials – Scientific guideline | European Medicines Agency (EMA)
9. Acetaminophen Use in Pregnancy and Neurodevelopmental Outcomes | ACOG
10. AI Drug Safety in Pregnancy | Pregnancy | JAMA | JAMA Network