Closing the gender health gap: How biopharma R&D can take action

Early in my career as a neuroscientist studying epilepsy, I volunteered for a charity, staffing a helpline for newly diagnosed patients and their families. I quickly saw the impact of the condition on daily life – affecting work, social activities and, strikingly, women who wanted to start a family.

At the time, many anti-seizure medicines, especially older ones, carried a high risk of birth defects. Some practitioners advised women with epilepsy not to have children at all, despite evidence that with the right treatment plans, many could safely do so. That experience left a lasting impression and showed me how oversimplified approaches and systemic blind spots can create serious care gaps.

Today, the structural and systemic drivers of health disparities between women and men are well documented, including by the World Economic Forum. In biopharma research and development (R&D), we have an unprecedented opportunity to help close these gaps, powered by recent advances in science, technology and data.

To deliver truly innovative, impactful medicines, we must build meaningful patient representation into every stage of R&D, from preclinical research through clinical trials to post-market studies. The question is: what does closing the gender health gap look like in practice?

Women’s health needs extend far beyond pregnancy and childbirth. Many diseases disproportionately affect women, particularly autoimmune conditions, which impact an estimated 230 million people worldwide and are the third most common disease category in the US after cancer and heart disease.

Women account for about 80% of autoimmune patients overall, with some striking disparities: roughly 90% of lupus patients are women, women are up to 16 times more likely than men to develop Sjögren’s disease, and twice as likely to have chronic spontaneous urticaria (chronic hives).

These differences highlight the need to account for gender-specific factors in drug discovery and development. Our company approaches this challenge by first building a deep understanding of disease at every level, from atomic and molecular to patients and systems. This includes prioritizing representation at the earliest stages of preclinical research; for example, selecting preclinical models that minimize the risk of generating confounding data in diseases that disproportionately affect women.

Importantly, we’re working across our industry to improve awareness and adoption of best practices. We’re committed to advocating for more representative preclinical research, better animal and cell models, greater attention to genetic variations, better data and algorithms designed to reduce data bias, and more. Applied broadly, these principles can help our industry develop safer, more effective medicines that improve outcomes and quality of life for all patient groups.

Data and AI offer a powerful opportunity to close the gender health gap. However, if sex differences are not accurately captured and reflected in health data, AI models may perpetuate bias and deepen disparities.

To address this, biopharma companies are investing in robust data platforms. One of our company’s key initiatives is Data42, an internal data lake that brings together decades of Novartis preclinical and clinical data into a single, accessible resource. This AI-enabled “treasure trove” includes annotations for sex, age, family history, disease conditions and more. It spans thousands of clinical trials with data representing more than a million anonymized patients, plus real-world evidence from hundreds of millions more, allowing teams to efficiently test scientific hypotheses.

These types of resources help us build AI models that better account for sex-related differences. In one example, an external preclinical study suggested a common drug reduced neuroinflammation in female animal models. Using Data42, we examined whether it might provide greater benefit for women with a certain neurological disorder.

While real-world and clinical data confirmed that the disorder affects women about three times more often than men, we saw no sex-specific difference in the drug’s potential clinical impact. This data-enabled insight suggested that the preclinical signal might not translate to humans and helped us redirect efforts to other projects.

Once a drug candidate reaches human testing, representation in clinical trials is critical. Sex-based differences should be proactively studied, including pregnancy and breastfeeding. This reduces blind spots that can lead to suboptimal efficacy or disproportionate side effects.

Recent guidance from the International Council for Harmonisation stresses that participant selection should mirror the groups a medicine is intended to benefit, so results are generalizable. This is especially important for conditions that disproportionately affect women. For example, in a recent phase 3 programme we conducted in chronic spontaneous urticaria, we enrolled female participants in proportion to real-world disease prevalence in the US, which helped improve our understanding of safety and efficacy.

More broadly, we are working to make all phases of clinical research more inclusive by strengthening partnerships with patient organizations, medical institutions and community groups; exploring alternative recruitment models and innovative study designs; and contributing to new guidance and standards.

R&D also continues after trials conclude. Closing health gaps requires ongoing learning from patients. We analyse endpoints, collect trial design feedback, capture lived experiences, and more. Even seemingly minor details can be important. Take wearable devices, for example, such as chest monitors used in clinical trials for ECG measurements. Considerations like comfort and wearability for people with different chest shapes may potentially impact the quality of data generated.

We can capture these kinds of insights through programmes such as our internal, AI-enabled Patient Insights Navigator, to ensure researchers have streamlined access to information that could help improve future trial designs. Together with Data42 and other resources, this allows us to build institutional knowledge on what matters most to patients, and helps inform and dismantle barriers to enrolment and retention for women and other underrepresented patient populations.

Post-marketing studies and safety monitoring are equally important, providing real-world evidence that can reveal differences across subpopulations and steer future R&D. Our PRegnancy outcomes Intensive Monitoring Program (PRIM), for instance, is an initiative that collects pregnancy data to offer timely evidence for healthcare professionals, patients and scientists working on the next generation of medicines.

Addressing the gender health gap is not optional; it is fundamental to achieving better outcomes and a healthier future. Closing this gap will transform care for billions of women and unlock enormous societal value, with the McKinsey Health Institute estimating an economic impact of up to $1 trillion a year by 2040.

The decisions we make now across biopharmaceutical R&D – what we study, whom we include, and how we design and test our medicines – are already defining the future of women’s health. This is how we take action.