Tackling the gender health gap with user-centred product development

Disaggregated data – designing medical products based on different user needs

It is important to be aware of biases in existing data. Clinical trials are starting to include more women, but this should be more than a box-ticking exercise. If this is not done proportionately and not followed up with sex-disaggregated data analysis, in other words, acknowledging and studying sex-specific results, we will continue to see health gaps and create new ones. There could be different outcomes and experiences between sexes that we don’t necessarily predict. Without systematically looking for them in our analysis, we won’t find them and design for all users of the device. 

This does not only apply to sex-disaggregation – there are other gaps too. If research and recruitment systems are not built to actively include and understand women from different socioeconomic backgrounds, ages, identities, geographies, races and ethnicities, disparities that marginalised groups already face will be widened. 

The same is true in the design of medical technologies, where you need to ask at every turn: could there be a different experience, biology, or context of use that I am not acknowledging? Without answering this question through inclusive research and analysis, the requirements defined may be based on assumptions that only apply to certain groups.  For example, the accuracy of pulse oximetry devices is decreased in people with darker skin, and pulse co-oximetry devices can overestimate haemoglobin readings in pregnant women.  

Innovating in context – addressing systemic bias

Medical technologies must be designed with an awareness of systemic bias. You can’t innovate in a vacuum. Whatever the technology, it’s going to be implemented in a wider system and used by people who may have and/or face subconscious biases. The more we can learn about these biases up front, the more we can address them in our approaches to research and in the design of medical technologies.

For example, consider that in the current women’s healthcare pathways, many tests, treatments and interventions (e.g. pessary provisions and other gynaecological services) are accessed in specialised care settings, often with long waiting times. Referrals to secondary care and setting up multiple appointments in different clinics and hospitals presents a significant access barrier for people who don’t have the time or resources. This means that if we focus on data only from these secondary care settings, it may suffer from sampling bias. Once more, there is a clear need for a systematic approach to collecting as broad a range of data as possible, to ensure the FemTech solutions we create are designed for everyone.

Learning (carefully) from history

In the medical industry we are standing on the shoulders of giants. We build upon existing tools, systems and techniques and find innovative ways to improve them. The challenge is that when carrying over the efficiencies, innovation and successes of existing devices, we are also perpetuating the disparities and biases that may have been inadvertently built in before. 

To tackle this, we consciously boil things down to the most basic of needs and requirements across population demographics and acknowledge our own “default bias”. We always have a frame of reference that influences our thinking, often sub-consciously. As device designers, we often think of our personal experiences and the devices we’ve seen before. This can help us find pain-points to tackle, but it can also inadvertently dampen creativity and perpetuate health gaps. 

When building devices that improve and draw upon existing procedures, it is important to ask yourself – is this the best solution for all users? Does a procedure have to be invasive? Does it have to be used in this clinical context? For example, the number of women screened for cervical cancer could increase by 400,000 per year with the adoption of at-home screening kits.  

Does it cause pain or discomfort and if so, what can be done to prevent that? What are the origins of this procedure? Is it leveraging tools and devices that have been repurposed for women, but not designed for them? For example, laparoscopic devices can be ill-suited for use by female surgeons, creating ergonomic injury risk.  

The first artificial hearts were too large for women, and women have been found to have more complications from procedures with devices like stents and left ventricular assistive devices. Because of the lack of medical research with female participants (the FDA and NIH did not mandate inclusion of women until 1993) and lack of sex-disaggregation (a 2014 study found that only 14% of post-approval medical device studies included sex as a key outcome measure) there may be devices out there that work differently for women without our knowledge. The consideration of anatomical and ergonomic differences across patient groups needs to be built into the device requirements to prevent these disparities.