What is the Gender Data Health Gap?

The gender data health gap is the difference in quality and quantity of health data both collected and analyzed between women and men. 

Why does the Gender Data Health Gap matter?

The gender data health gap is often responsible (whether on a conscious or unconscious level) for the perpetuation of disparity in healthcare that women and men receive. This can result in the following: 

Delays in diagnosis: Women wait an average of 4 years longer to receive a diagnosis for the same disease as men.

Many of the ways in which we diagnose diseases (e.g. ‘cut-off’ values, symptoms or even the instruments used) are overly reliant on the white male model of presenting, which results in women being underdiagnosed for diseases that are present in both sexes but manifest differently. 

Not being prescribed the appropriate treatment:
p.ex. In a cohort study of adults with acute nontraumatic abdominal pain, women were 13% – 25% LESS likely to receive opioids in the emergency room for their pain despite presenting with the same pain scores. 

Conversely, In the UK, women in England were 59% more likely to be prescribed benzodiazepines (medication often used to combat anxiety and insomnia)– better known by the brand names of Valium, Xanax and Temazapam than men between January 2017 and December 2021. In 2020, the FDA mandated that a “black box warning” be placed on benzodiazepines to inform patients that withdrawal from the drugs can be life-threatening. 

Care that is not structured to take differences in account:   

For instance, some research has indicated that women athletes are more susceptible to muscle and tendon injury during ovulation. An interesting example of tailoring training to changes in the menstrual cycle is: Wild.ai. Female athletes may also be more susceptible to having concurrent issues like eating disorders, multiple stress fractures, gastrointestinal issues and mental health concerns but the approach to these issues (although they are related) is fragmented, there is not yet an interdisciplinary approach to these issues. 

Who does it impact?

The Gender Data Health Gap is embedded in the workflow of each and every stakeholder in the healthcare ecosystem, whether they are conscious of it or not. Examples include (but are not limited to):  

→ A lack of training on how diseases may present differently in women or conditions that only impact women:

p.ex 41% of UK universities do not have mandatory menopause education on the curriculum, thought it is an integral transition all women will go through at some point in their lives.

Resulting in statistics like: 1 in 3 women between 45 to 54 being given an incorrect diagnosis before finding out their symptoms are related to menopause (and 32% of women feeling their doctor was not very knowledgeable about the topic). 

→ Structural bias in diagnostic tools:

p.ex Heart Attack:
Cardiac troponin (cTn) test: used to measure the level of troponin (protein released by damaged heart muscle) in the blood. Higher levels of troponin are used as an evaluation parameter for more heart damage. The clinical threshold that signals a heart attack can differ between men and women, i.e. a woman could be having a heart attack but the troponin level would be below the level of detection. 

Cardiac catheterization: used to detect blockages in large arteries. Women are more likely than men to have plaque buildup in the smallest arteries due to inflammation which could be better visualized with an MRI.

If the tools used for screening for heart attacks render ‘invisible’ the heart attacks women have, this can further contribute to misconceptions about what type of individual has a heart attack and is a part of a reason (but not the only one): women have a 50% greater chance of misdiagnosis of a heart attack compared to men.

Lack of treatment options to prescribe to patients:

p.ex: there are only 2 FDA-approved treatments for female sexual dysfunction (which impacts approx. 40% of women) in the world vs 27 treatment options for men. And the first treatment option for women was only approved in 2015 vs 1998 for men.

→ Patient dismissal:

84% of women report feeling dismissed by their GP in the UK. Nearly 1 in 4 women say they do not feel their clinician takes their pain seriously (versus 1 in 6 men). 

→ Lack of awareness around female-centric diseases, diseases that disproportionately impact women or diseases that present differently in women result in:

– delays in diagnosis (from the patient side) as women are not aware that they should seek care (or that care is indeed available) 

-an incomplete picture of potential complications of the condition. Eg the link between PCOS and cardiovascular disease (studies suggest that women with PCOS have 2x risk of a future cardiovascular event, like a heart attack or stroke). 

→ Lack of sex disaggregated data: 

p.ex. In COVID-19 vaccine trials, 28.3% of publication did not report sex distribution among participants and only 8.8% of the studies provided sex-disaggregated Vaccine Effectiveness estimates.

Sex-disaggregated data is essential to understand whether there are increased side-effects in one sex or the other and more generally to understand the distributions of risk, infection and disease in the population. For example, a study that looks at the impact of a new drug on pain relief might draw the overall conclusion that a drug is effective for both men and women.

But disaggregating would allow visibility into whether the drug works better in one sex than the other (NB women and men may well have different mechanisms for experiencing pain). Data disaggregated by sex also allows better visibility into how resources are allocated (according to a 2022 report by the National Institutes of Health (NIH), NIH funding for women’s health research was $6.3 billion, while funding for men’s health research was $9.2 billion: a gender funding gap of 39%)

→ Not including equal numbers of females and males in their studies: 


→ Not having visibility into where the gaps in women’s health research are: 

p.ex. The National Institutes of Health (NIH) has not assigned a unique identifier code to menopause, unlike other conditions such as anorexia or prostate cancer. This means that anyone who wants to know how much funding the NIH has awarded for research on menopause must manually count the number of grants that mention “menopause” in their titles or descriptions.


→ A lack of data into prevalence and incidence of disease:

The incompleteness and inaccuracy of women’s health data, as well as the lack of standardization in electronic health records, pose challenges for FemTech companies to gather and analyze large-scale data to identify specific targets for developing solutions that effectively address women’s needs.

→ ‘Invisibility’ of the issue they are targeting (resulting in difficulty monetizing and marketing):

How do you market a solution when there is a fundamental lack of data and research surrounding the problem space?
Lack of data makes it difficult to monetize a solution with larger institutions when those in charge may not experience the problem area the solution is solving for themselves. Data acts as a substitute for lived experience.

FemTech startups face an uphill battle due to often not only having to validate the efficacy of the solution they are bringing to market but convincing stakeholders that it is an issue they need a solution for in the first place.

→ Lack of understanding of the cause of many ‘common’ women’s health conditions makes it difficult to develop diagnostic and/or therapeutic solutions:

P.ex: Endometriosis (impacting approx. 1 in 10 women), Uterine fibroids (up to 77% of women during childbearing years), PCOS (1 in 10 women), PMDD (between 1-12 and 1 in 20 women of childbearing age).
The exact cause of all of these conditions is unknown – and consequently each have dissatisfactory means of diagnosis and treatment.

Why address this now?

Well, it’s 2023 – if not now, when?  But more specifically in the context of potential widespread AI adoption in healthcare, we run the risk of embedding these biases and gaps in a structural and systemic manner without even realizing it. 

AI learns from the data it’s trained on so if women are ‘invisible’ in that data or mis-represented this can have lethal consequences. 

On the flipside, it is worth noting that given how extensive and pervasive this gender data health gap, AI may well be the only way to close the gap in a relatively timely manner.

A few examples of how AI could be used in a beneficial manner:

AI can help to identify and address gender bias in health data. AI can be used to analyze large datasets of health data to identify patterns and trends that may be indicative of gender bias. This information can then be used to develop strategies to address the bias.

AI can help to collect more comprehensive data on women’s health. AI can be used to develop new methods for collecting data on women’s health, such as using wearable devices or social media data. This data can then be used to improve our understanding of women’s health and to develop more effective treatments.

AI can help to develop more personalized and targeted health interventions for women. AI can be used to analyze data on individual women’s health to identify their unique risk factors and needs. This information can then be used to develop personalized and targeted health interventions that are more likely to be effective.

TL;DR What does FemTechnology Summit have to do with all of this?

At this year’s FemTechnology Summit (held on June 7th 2023), stakeholders all across the women’s health ecosystem (FemTech Startups, clinicians, researchers, corporates in the pharmaceutical industry) were assembled to tackle the Gender Data Health Gap.

The objective of the workshop was to analyze commonalities or overlaps in the gender data health gaps in their workflows of different stakeholders to see if an overarching narrative about the gender data health gap could be established and see what the best initial steps towards closing the gap would be. 

What can we leverage to tackle the gender data health gap in real time?
Patient Reported Outcomes:

What is it women want to know about their own health, where are they looking for answers and consistently not finding solutions. How can we prioritise the problems women themselves want to be solved in order to identify the most pressing ‘unmet needs’. 

Some examples of how to execute? 

Clue, the #1 doctor-recommended free period tracker app built in collaboration with top health researchers, is a great case study in this. In the words of, Audrey Tsang, Co-Ceo, Clue: We often hear users in our community say that ‘I just want to be taken seriously’. The world today doesn’t take their pain or their concerns that ‘something doesn’t feel right’ seriously. That’s why they track in Clue—so that their data can help them advocate for themselves and the care they need” 

Or Roche’s #MyStoryForChange initiative: where the stories of 600 women across the globe were collected to better understand the bias in interaction with the medical system women experience. 

A theme that emerged time and time again in those stories was that in healthcare settings, many women feel like they’re not being listened to or that their experiences are not being taken seriously. 

There is a need to find alternative ways to access the healthcare experiences of women, to really learn what issues women are struggling with. It’s not that the information is not there – people are asking so many questions about their own health concerns. But if it’s not being captured, and harnessed, then does it even exist? 

Collecting novel biomarkers and datasets via FemTech Startups:

FemTech startups are uniquely poised to collect data sets that have been previously neglected. Some example of participants in this year’s Gender Data Health Gap Workshop doing just that are: 
Impli – continuous hormone monitoring
Daye – vaginal microbiome sequencing
TheBlood – analyzing menstrual blood for unique biomarkers in women’s health. 
Sanno – digital biomarkers for gastrointestinal issues

Data lays the groundwork for any  AI application. As we have seen women’s health is rife with data/information that have not yet been researched or collected. FemTech is uniquely poised to disrupt this by collecting novel women’s health data sets in real time.  

Reimagining care:

Women are more likely to suffer from chronic conditions yet our current healthcare model is structured to service acute situations (and is episodic in how it engages patients). 

If we employed a more consistent, longitudinal means of engaging patients that would expand the type of information we are able to collect (eg noticing patterns in depressive episodes that might be linked to hormonal fluctuations).  


It is more vital than ever to remember that closing the gender data health gap requires action from all stakeholders. We need to weave all this data together: we at last have the technology needed to collect this information at-scale. It cannot exist in siloes. We must sure that innovation existing in the domain of one stakeholder (it is estimated that it currently takes an average of 17 years for research to be translated to the bedside, and although we know that patients frequently share their concerns in patient forums that data is seldom translated into practice) is able to be seamlessly implemented into practice by other stakeholders.   

We can no longer afford to silo reproductive health from “general health” (aka other organ systems) when we know that reproductive transitions are often critical phases where protective action for eg cardiovascular health can be taken.  

Healthcare is currently being redesigned. It has to be – AI will undoubtedly play a pivotal role in that redesign and as such presents an interesting opportunity. However, if our current healthcare model has taught us anything it is that we want the structure to be sound (aka have the right data sets to base these models off). It is much harder to undo something that is flawed at its core and be forced to build around it with patchwork pieces.

We need to accelerate the collection of these data sets now. 


TL; DR What can I do?
Are you a patient experiencing these gaps in care? 
Are you a clinician who has a perspective to contribute about what tools you feel are missing for you to adequately serve your patients? 

Are you a researcher? And have been studying novel links between different women’s health conditions or reproductive health and ‘general health’? We’d love to learn more about it!

Are you a start-up who is interested in collaborating/conducting further research on novel data sets you have collected.  

Contact form 1