Invisible unknowns: How the Health Tech Hive studied women’s health

x Sponsored content on Research Live and in Impact magazine is editorially independent.
Find out more about advertising and sponsorship.

Women are 50 per cent more likely to be misdiagnosed after a heart attack because they have different symptoms to men, while, on average, they receive a diagnosis for diabetes 4.5 years later. For women suffering from endometriosis, it can take an average of 7.5 years to be diagnosed.

There are still glaring inequalities when it comes to the health of women – in how it is understood, discussed and treated. Caroline Criado Perez’s book Invisible Women, published in 2019, laid bare the problems caused by a lack of sex-disaggregated data, from heart attacks being misdiagnosed to car-safety systems not taking account of women’s measurements.

It is one thing to recognise that an area of health needs more research, investment, solutions and awareness, but another to identify exactly what the needs are within that. This was the issue facing the Health Tech Hive, a network of policy-makers, entrepreneurs, engineers, academics and competitors, set up with the aim of building technology solutions to health challenges.

Established in response to what co-founders Jasmine Eskenzi and Brigitte West saw as an increasingly siloed health-technology space, the group focused its attention on women’s health as a key challenge because of a general awareness that there was a “massive data gap”, says Eskenzi. “Men’s bodies are seen as the status quo. It’s time we changed this.”

While the existence of a data gap was known, however, the scale of the gap – and the conditions and issues affecting women as a result of it – were unclear. So, the Health Tech Hive partnered with Discover.ai to conduct research on the topic.

They used artificial intelligence to undertake a qualitative analysis of a range of online sources of conversations between women, including forums, social media, advice columns and blogs. The research combined language from 251 sources across 19 sourcing areas and two markets – the UK and the US. It aimed to uncover cultural insights and stories that would examine the unmet needs in the area of women’s health, and identify target areas for future research and strategies.

An academic literature review would not have worked for the purposes of identifying what the invisible needs were, for the simple reason that there is not enough existing research. “You’re trying to work out what people haven’t been doing research on,” says Hannah Marcus, strategist at Discover.ai.

“We thought that by going to these online spaces where people are choosing to talk about what’s happening to them – that might be in a health space or an adjacent space – we’d get spontaneous and unmoderated stories and voices across different online spaces, and see how we could bring them together.”

The analysis drew insight from some surprising places. “Goodreads is an interesting source of data. If you’ve got a book talking about a data gap, such as Invisible Women, a lot of people’s reviews will talk about why it resonated with them and their experiences. Suddenly, you’ve got this spontaneous response to a text, and we tried to capture a lot of it.”

For a subject matter as broad as women’s health, managing the potential scale of the project was challenging, says Marcus. There were multiple possible areas on which to focus, and the research needed to be as comprehensive as possible without being overwhelming.

“We had some quite interesting conversations about how much we should talk about cancer, for example,” says Marcus. “That’s obviously a very big area, but also one that has research and funding. It’s not as completely unresearched in the same ways.”

Birth control was another topic that came under scrutiny, says Marcus. “That is a real area of tech innovation – but not a condition in and of itself.”

In the end, the research had two areas of focus. First, the needs specifically affecting people with women’s bodies, including conditions such as endometriosis, perimenopause, menopause and polycystic ovary syndrome, as well as issues such as fertility – including miscarriage – vulvic pain, thrush and sexual pleasure.

The research also flagged conditions that affect everyone, but for which the impact on women is under-researched or unknown: noncommunicable or chronic diseases (NCDs); heart disease; osteoporosis and joint conditions; auto-immune diseases; diabetes; weight; incontinence; cognitive decline; Covid-19 and mental health.

Discover.ai mapped these issues against what it termed ‘macro causes’ – eight hypotheses for why gaps exist in data. These ranged from women being disbelieved by medical professionals and the way in which pain is considered in women, to the assumption that women are ‘small men’ and that any data already held is sufficient.

Following the research, the Health Tech Hive is working with various investment groups to assess the current gaps in investment and identify where quality is lacking, in terms of health technology, around the various health needs.

“It might be that there is no funding going into a particular area, or that there is a massive gap in the academic work,” says Eskenzi.

The initiative also hopes to work with more partners to create working groups around each of the areas identified within the research. Eskenzi was struck by the scale of the issue.

“We had known this was a big problem, but this was not a clinical tick-box study; this was going into forums and [gathering data on] what is being spoken candidly about among women.

“There is just so much we need to do. We really need to take this research and scream it from the rooftops that there is a massive issue here. We need to put funding behind it; we need to put minds behind it – we need to address it.

“Women have been dismissed by so many doctors, been told ‘it’s part of life’. It’s been far too long that women haven’t been heard.”

This article was first published in the January 2021 issue of Impact.