Addressing the Gendered Health Data Gap

When thinking about personalised treatment today, we think of targeted cancer treatments, or gene therapies. Treatments that are designed for small subgroups of diseases or people and which are more effective because they are designed specifically for what they intend to do.

Whilst we have made huge strides in improving treatment for some of these conditions, it seems we are still missing a trick. Without the need for technological advances or medical innovation, we could better design our health systems and our treatments for 50% of the patient population. It seems an obvious place to start.

Everything created in our world has been thought through in terms of design: how high to put the shelving units in the supermarket, the temperature of your office and the safety specifications of your car. In making these decisions engineers and designers look at data to understand the optimal approach.

However, this only works when the data is representative of the people the design is aiming to serve, and all too often it isn’t. Women are often underrepresented or missing from these data sets resulting in a world that has been designed for men as default. This is also true of our healthcare systems; our clinical trials, our diagnostic approaches and our treatment options, all typically have men as the default.

Take heart attacks for example, the British Heart Foundation found women are 50% more likely to be misdiagnosed following a heart attack. One of the routine tests to diagnose a heart attack is the troponin test used to identify damage to heart cells. However, it was found that the expected ‘normal’ limits set were those that were normal for men as the default, leading to a higher incidence of women’s heart attacks going misdiagnosed. Women were missing from the data sets used to build the limits, and therefore the test becomes much less effective for women.

By using a more sensitive test and altering the expected limits to be appropriate for women, the number of women diagnosed with a heart attack doubled – women who otherwise would not have received a diagnosis, leading to delay in treatment and therefore worse outcomes overall.

As Caroline Cirado Perez points out in her book Invisible Women, this isn’t some malicious attempt to exclude half of the population from the world we design, it is just a consequence of non-representation. By not including women in the data sets used we continue to build systems with men as the default and this can have serious, and potentially deadly, consequences.

However, as the example above shows, the potential for positive change can be huge. By simply splitting data sets by sex, we can quickly and easily identify missing data, and find opportunities where addressing these could make a big impact to addressing gendered health inequalities. If this had been done in setting heart attack limits it would have become clear that there should be different levels for men and women.

The WHO’s Global Health Statistics were disaggregated by sex for the first time in 2019. So, whilst it’s a wonder that this wasn’t considered before 2019, it’s a step in the right direction. It’s a starting line.

In the UK, the Government are looking to address gendered health inequalities, bringing forward England’s first Women’s Health Strategy. Putting aside the fact that even the need for this clearly shows that women’s health isn’t covered by ‘health’ in itself, it is a great opportunity to address the gendered health data gap.

As we start to put more and more faith in data, the responsibility to make sure this data is representative of both women and men becomes stronger. A key part of the Women’s Health Strategy should be on ensuring women are included in the data sets used in health system design, clinical trials and disease management and also that the data is disaggregated to identify gendered differences.

Disaggregating these data has clear benefits for the female 50% of the population but it also has benefits for other groups; race, ethnicity and socioeconomic factors should all be taken into account too.

Not only will this ensure we are designing systems for all the people that will use them, but it will also mean we are able to see solutions we wouldn’t otherwise and uncover things that don’t intuitively feel true, or that we or others would never think to consider. For example, despite pregnant women being at increased risk from COVID-19, it took a campaigning group for them to gain access to the vaccine, a result that benefits the health system as a whole. At a time when the health system is recovering from the COVID-19 pandemic, finding smart, effective solutions to improving the health of the nation is paramount.

It is fascinating to think of the breakthroughs we may be able to take in medicine simply by challenging ourselves to ask the right questions and look at the right data. All of us are designers in our everyday lives, whether designing disease awareness campaigns or creating policy asks. We can all take a better look at the insights and data we use to inform our decisions and design, but we should also feel buoyed by the impact that we can have in making the world one that works for everyone.