Health systems – all around the world and no matter how advanced – are missing critical data.
Data is essential to the modern world. It drives everything - from innovation and research, to business decisions and importantly, healthcare. But healthcare systems have largely been built around data collected - be that through clinical trials or innovative technologies - and shaped, by men. For too long, women have been systemically overlooked and under-represented in healthcare, causing a gender data gap that has led to a lack of knowledge about conditions that primarily or predominantly affect women, or affect women differently.
Women’s health is often seen narrowly, in terms of specific body parts, and many conditions that affect both men and women aren’t studied through a sex-specific lens. For example, up until 1993, women were excluded from clinical research and trials.1 While this has improved in recent years, women are still often underrepresented in many parts of the clinical research process, and even when women are included, many studies still do not analyse data by sex.2
Gender inequities in access to innovative technologies - such as artificial intelligence (AI), machine learning and mobile technology - designed to collect health data, can also fuel the gender data gap. This can be further influenced by differences in education, income, age, ethnicity, and location.3 This inequity can be particularly impactful with AI systems, which replicate patterns of gender bias in ways that can exacerbate the current gender divide.
These gaps in data have led to a fundamental lack of understanding of the influences of sex on the prevalence, presentation, and progression of dieases,2 which has a number of resulting health inequalities. For example, a lack of understanding of the differing symptoms between sexes means the risk of misdiagnosis for a heart attack in women is 50% higher than in men. Combined with the fact that women are less likely to be prescribed medicines to reduce their risk of suffering another heart attack, women are more likely than men to die after a heart emergency.4
At Roche, we believe research and clinical development require greater diversity and representation of the people who experience a disease, and we understand the value of diversifying genetic data for scientific discovery. It’s time for change, but we know that we cannot address the gender data gap alone, and there is work to be done to close the gap in data gathered from clinical research as well as from digital technologies. That’s why Roche brought together students from a variety of backgrounds and fields from around the world to tackle this issue in the
The InCube participants worked together to consider how we can address the data bias in AI and digital technologies, and account for sex and gender differences in their design. “There is so much potential when it comes to digital technologies for driving better data collection for women’s health,” said one of the participants, electrical engineering master’s student, Nektarios Totikos. “Be it an app-based symptoms tracker, using language processing to extract meaningful health data from social media, or implementation of fairness algorithms in AI models to better pre-process data on conditions that impact both men and women.” But in order to find a solution, we first have to recognise the issue. “Through this experience I was made aware of the scale of the gender data bias in healthcare, and the importance of collaborating to develop innovative ideas to remedy this disparity,” said Nektarios. “So much of what we discussed during the challenge led back to the fact that people don’t appreciate there is an issue.”
When stakeholders come together, we can make a difference and generate the data that are needed to close this gap - but we need to go back to the start. In this series of four articles, we're diving into the data gap to explore its societal impact, and how we are already triggering change by ensuring women are fully represented at every stage of the process, from research and leadership, to the use of new technologies.
Mauvais-Jarvis F, et al. Sex and gender: modifiers of health, disease, and medicine. Lancet. 2020 Aug 22;396(10250):565-582. DOI: 10.1016/S0140-6736(20)31561-0.
Jianhua W, et al. Editor’s Choice - Impact of initial hospital diagnosis on mortality for acute myocardial infarction: A national cohort study, European Heart Journal. Acute Cardiovascular Care. 2018 March. 7; 2:139–148. DOI: