Historically, women have been underrepresented in clinical trials due to concerns over potential impacts on reproductive health or due to perceived hormonal variability complicating study results. This has resulted in a critical lack of data on how conditions manifest differently in women or how pharmaceuticals behave in female physiology. Consequently, medical devices are often designed for male anatomy, and drug dosages are typically optimised for male subjects, leading to poorer health outcomes and a higher incidence of adverse drug reactions (ADRs) in women.
The historical exclusion of women from clinical trials has created a pervasive gender gap in medical knowledge, leading to a healthcare system built around the “male as default” model. However, the advent of real-world data (RWD), real-world evidence (RWE), and advanced artificial intelligence (AI) provides an unprecedented opportunity to address these systemic issues, dramatically improving patient safety, increasing access to appropriate medicines, and closing the long-standing equity gap for women.
How History Created Medicine’s Gender Gap
Women’s exclusion from clinical research didn’t happen by accident. Beginning in the 1950s and 1960s, tragedies such as thalidomide prompted strict protective rules that kept pregnant women – and often all women of childbearing age – out of early-phase clinical trials. Add in the misconception that female hormonal cycles made women “too variable” for research, and entire generations of medical evidence were built without women in mind.
This legacy and the complex question of protecting women’s health and balancing the risk for the unborn baby has resulted in the most significant gap in medical research today: pregnant and lactating women, who are almost universally excluded from clinical trials. When these women become ill, clinicians must often rely on limited data or extrapolate from non-pregnant populations, creating significant ethical dilemmas and potential risks.
Why RWE and AI Are Changing the Game
RWE and AI are providing the tools needed to rewrite the evidence base. Large-scale RWD from electronic healthcare records (EHRs), claims, and wearables reveals how diseases actually present in women. AI can sift through these patterns to uncover sex-specific symptoms that traditional trials missed.
Cardiovascular disease is a powerful example: many women never experience the “classic” chest pain associated with heart attacks. RWE helped surface this insight at scale, prompting changes in guidelines and clinical education.
AI and RWE can help to ensure safer medicines and appropriate dosing as AI-driven analysis of RWD allows earlier detection of sex-specific safety signals and supports dosing adjustments tailored to female physiology. Moreover, machine learning models can integrate hormonal status, genetics, reproductive stage, and comorbidities to guide individualised care for every woman—moving beyond the one-size-fits-all approach that has persisted for far too long.
Filling the Pregnancy Evidence Gap
Perhaps the area in women’s health where RWE can be most transformative is pregnancy. Observational RWE allows researchers to study medications already used in pregnant patients without exposing them to additional risk.
AI can detect patterns, highlight safety signals, and help regulators make clearer, more confident recommendations. This is a leap forward for maternal safety and access.
Safeguarding the use of patients’ data to benefit all:
These tools are only as good as the data and governance behind them. With proper design, we must ensure that the new technologies are used carefully to avoid potential flaws that already exist, examples include:
- Historical Data Bias: RWE reflects real-world care — which has not treated women equitably.
- Algorithmic Bias: AI models trained on biased datasets can reinforce the very gaps we’re trying to close.
- Missing Data: Pregnancy, menopause, and certain women’s conditions remain poorly captured in many RWD sources.
- Opaque Models: Black-box algorithms can erode trust and hide inequitable outcomes.
RWE and AI must be deployed with transparency and oversight. Research consortiums like Real4Reg where multi-stakeholder alliances spanning regulators, academia, industry, clinicians, and patients work together play a transformational role. Real4Reg, for example, investigates the complexity of sex differences in drug response at all stages across the product lifecycle and puts a special focus on patient groups that are often neglected in clinical trials such as pregnant women in some analyses. By involving patient representatives and regulators from the outset, Real4Reg and other consortiums ensure AI systems are explainable, equitable, audited and trusted by both clinicians and patients.
RWE and AI give us the tools to finally move beyond the outdated “male-as-default” model in healthcare. RWD collection initiatives like the European Health Data Space and more specifically for regulators, Darwin EU will give us the opportunity to see potential gender biases in outcomes from the data. This will offer a path to safer medicines, smarter diagnostics, and equitable care for every woman—pregnant or not, young or old, healthy or chronically ill. The evidence gap didn’t appear overnight, but with the right data, technology, and collaboration, we can close it far faster than it was created.