Hertility CEO Helen O’Neill on building a foundational model for women’s health

0:00Hello, as usual, it's your host, Amy Lewin. Joining me today is Helen O 'Neill, CEO and co-founder of Hertility, the women's health startup which offers hormone testing and supports women through their fertility journeys. She's also an associate professor in reproductive and molecular genetics at University College London. And as you'll soon discover listening to this podcast, one of the most engaging and informed people to listen to on the subject of women's health. Hertility, which Helen launched seven years ago with her twin sister Deidre and Natalie Gatru, has now raised over$29 million from investors, including Local Globe, Vemrex, West River Group and the actress Emma Watson, and has completed over 700 ,000 health assessments and taken over 100 ,000 bloods.

0:45Helen, it's such a treat to have you on the show. Good to see you. So Helen, what compelled you, and for anyone who isn't British, I am being sarcastic here, to combine the nice, stable life of an academic with the nice, stable life of a founder and launch this company? I think in retrospect, it's that classic thing where they say you would never do something if you knew. But I do think it was definitely something that I felt compelled to do. And I know that that comes from working to increase your knowledge on a subject, to know everything about a subject, and then being met with kind of a brick wall in terms of being able to translate that into action.

1:34And so I've spent the last 19 years in the world of women's health, early embryology, genetics, and doing research on all of the above. And when you work in women's health, it really invites conversations with women wherever you are. And when those conversations are always met with a lack of ability to translate it into health, despite your knowledge, then you just feel frustrated that you think, wow, as a molecular geneticist, I know that we now have cures for some of the rarest of genetic conditions. And yet we don't even have diagnoses or treatments for some of the most prevalent of women's health conditions.

2:19So I kind of felt enough was enough. And what's the first problem Hattility set about tackling? So we actually began as a clinical trial. And so that clinical trial was based off a grant that I got to develop predictive diagnostic algorithms for gynecological conditions. And this was really with the knowledge that so many symptoms that women face are quite insidious, whether it's pain or bleeding or weight gain, weight loss, hair loss, all of these different symptoms that can be mistaken for many different other pathologies or conditions. and the fact that pain tends to be normalized in women as well.

3:03So we really set out to see if we can pull apart the difference in how we treat and diagnose women. In the first instance, that really just looked at all of the known diagnostic criteria that we use that is fine, but clearly not doing a good enough job if average time to diagnosis in many of these conditions that affect one in 10 is up to nine years. and then trying to say, how can we use larger scale data to address one of the biggest problems that exists in, I guess, modern society now in terms of birth rates. So the first question we wanted to ask was, can we predict infertility? And in trying to answer that question, we had to get to the bottom of all of the things that could cause infertility.

3:49And that was really first looking at all of the gynecological conditions, which affect one in three women. And given that infertility now affects one in six, it's a massive problem. And why, it feels like fertility tests for men have been around a lot longer, like as a kind of over the, maybe not over the counter, but easily accessible to people. Why was women's health not so easy to look into? you? Women's health has always been considered, well first of all we've always considered health to be standard right, we use the male body as a reference point which I will tell you all about how wrong that is but the male body has essentially been used as our gold standard or our control group and that is because and this has been described that women are hormonally noisy and And what that means is that in every given month, our hormones are fluctuating quite significantly throughout the menstrual cycle.

4:52And this is true even of animal studies. So when even doing preclinical trials, they're typically done on male mice or they're not, they don't annotate what gender the animal model is, which is problematic as well. So the reason we know so little about women's health is because it's actually technically very difficult to work with women, our female bodies, especially with regard to reproduction. So this hormonal noise, while now people complain and they say women weren't included in clinical trials till 1993, and I get a little bit sick of this stat, even though it's very shocking. Partly it's because we have this blind spot in terms of acknowledging the sexual dimorphism all of our tissues of our beings and how we react to things.

5:40But the major reason is that it is quite difficult to collect blood from a woman at a single time point. So in order to collect reliable, robust data from a woman and keep it at the same time, it has to be done on the third day of her menstrual cycle. So a quick bit of calculation lets you know that even if you have regular cycles, you only get 12 of those opportunities every single year. And so collecting data from women on the third day of their menstrual cycle at scale is actually feasibly and technically very difficult to do, which is why women have never been included in clinical trials. But that's massively problematic, right?

6:19If you think that eight out of 10 drugs that were taken off the market were to do with adverse reactions in women, you start to see this picture of why it's so wrong that we don't A, include women, B, stratify data according to sex. And so us being able to overcome that difficulty in terms of our data, in terms of being able to overcome the barrier to day three testing by providing remote blood testing has actually meant that our data set is hugely valuable in terms of the medical sphere. But it's sometimes easy to complain about a problem about why women's health hasn't been given the attention it deserves, but there are logistical problems.

6:56No excuses for it, but there are logistical problems with it. So with this very valuable data set that you're building, what's the kind of long term vision for that? Like what can you do with that? Well, as I alluded to, given that the data doesn't exist, it is so valuable to have this data in terms of our ability to understand pathologies, diagnose pathologies that affect so many women and create drugs. we have been working to build this data set from a multimodal perspective. So we've got these health assessments, we've got blood results, and then we've got ultrasound images. And we've trained machine learning models and now AI to diagnose these conditions with 98 to 99 % confidence.

7:42So to me, when we look at the Women's Health landscape, it is crippling to see just how long it takes a woman to get to diagnosis and how much of a backlog there is in the current health system. So the creating a bedrock of data that is labeled, annotated patient data with known outcomes is hugely valuable for building further foundational models with regard to diagnosing and treating women and then building upon that for a multi-omic data set to really then answer some of the questions that have never been answered with regard to every reproductive life stage in women, but equally every condition.

8:25And so is that something you would partner with a healthcare system on? Is that something you will use to turn into fertility products? Is that something you will use in an academic sort of way? Like where do you take that? It's probably a good idea to take a step back and say how the current journey works and what we've built to date. So to date, we have built a very robust health assessment, which we have a patent pending on this health assessment. This health assessment enables us to triage according to different suspected pathologies. And that health assessment actually is one of, you know, 20 health assessments, depending on who you are, what your age is, what your intention is, and potentially any pathology you might have.

9:11So we have internally embedded algorithms within that health assessment that help us to determine your relative risk of a given pathology should you have one. The next step is to send you a blood test that you could do from home on the third day of your menstrual cycle to confirm any suspected pathology. So that's like quite a dynamic biomarker selection so that if you've triggered any suspected diagnosis and we can confirm that diagnosis with a set of analytes, we'll do that. And then the third step of that is interpretation of those bloods. So typically in a health system or any blood testing companies, they send people reams of results that will say, you know, your FSH is 8.9 picomoles per millilitre at a reference range of 8 to 100.

9:55And you look for this asterisk and you say, I'm out of range. What does this mean for me? And so we really worked hard to not just provide the blood results in a really understandable way, but to interpret them in the clinical context of the other blood results and all of the health assessment details that you've told us. This was a huge piece of work in terms of coding each of the different data points in the health assessment, whether it was a lifestyle factor like smoking or sleep patterns or whether it was a symptom or whether it was a menstrual factor. So that we coded over 54 ,000 clinical scenarios so that every report will say, hi, you've told us X, Y and Z and we've interpreted all of that for you in a in a consumer facing report as opposed to something that's intended for another clinician to decipher.

10:43And then we provide all of the requisite onward clinical care, whether it's scans, ultrasounds, whether it's IVF, egg freezing, telemedicine appointments, gynecologist prescriptions, etc. So you see that this kind of like clinical chameleon that we've built end to end enables us to then really use that as a loop to confirm diagnoses. and so that the first version of what we built really has been tested on different types of I guess buyers to a certain extent I've never really described it as that but it is probably useful to do because we first launched direct consumer so any woman could buy this and then we launched b2b so that employers are paying for this as a benefit and then we launched as paid for by insurance.

11:29So we've now tested this across three different types of systems, which has been really valuable in looking at different attitudes, but equally the different behaviors of that end user. It also means you get a different variety of, I guess, you get population level data, as opposed to sometimes you'll get a skewed data set towards people who are solely health conscious, or they're just doing it because their employer has paid. um that's basically been building something that is very much serving the customer or the patient for the last two years we've been building internally and that is people say like you can sell a million um of these health journeys but you can't scale a doctor and so um that's what we've been doing we've been scaling the doctor through a digital mean and it's amazing to see I guess this evolution of AI and know fully that I feel safe in the confidence that nobody has the data that we have for women's health and you can use AI and it will not diagnose you for women's health conditions because the data is not openly available and so we've built our own internal system which we call Gyn AI which is built on our own proprietary labeled and patient annotated data and that is the one that can predict and diagnose with 98 to 99 percent confidence and precision and recall for 25 different health states.

12:51And that guy in AI is helping clinicians. So the next version of, I guess, fertility that we will scale is that software as a medical device, which will be in the hands of primary care providers and secondary care providers. And what that does is it automates the clinical workflow. It automates the letter writing. It automates the interpretation of the bloods. And then it automates the ultrasonography. So we've built a model that can detect uterus, ovaries, fallopian tube and then write the letter for the clinician. So really scaling the clinician and reducing that unnecessary clinical workflow that is really not a good use of a clinician's time.

13:29So it means that they're having the opportunity to have meaningful discussions with their patients about what the next steps are for them. Okay, heaps to get into there. What would you say when you talked about so you've got the you've had the director consumer you've got the um people getting this through their health insurance and people getting this through work what are the what are the limitations of that i'm i'm broadly you know stereotyping here but i'm guessing that still means that most of the women whose data you have collected are wealthier um you know in work of a certain age, maybe a bit older, maybe not from certain communities?

14:14What do you know of what you have and what is missing partly as a result of how you... Well, you can assume that, but unfortunately, we're launched in the UK and Ireland, whereby actually globally, healthcare is in a... women's health is in a really dire state. But as a result of COVID, where all clinics that were deemed non-essential were closed, meaning all gynecology clinics were closed there are over 700 ,000 women on a wait list to see a gynecologist in the UK and so it's not really a surprise that over 700 ,000 women have done our health assessment even if you go through the NHS you are waiting on average nine months to see a gynecologist which is ironic given if your problem might have been that you missed a period and nine months later a baby arrives and you say you know what my problem has solved itself.

15:06No we don't just see wealthier older women 25 of that database are 18 to 25 year olds they're health conscious but they're also really experiencing symptoms and I sometimes equate it to so I equate it to the cost of an iPhone and yet a lot of people have an iPhone or they have the latest trainers people don't prioritize their health necessarily until it's urgent and our price point has never been one that excludes people we also use you know buy now pay later options like farina and buy now pay later um so no we don't actually have just um you know wealthy white women we have asked for over 80 um different ethnicities and have a hugely diverse data set in fact our data set is as diverse as the uk population so it's it's almost mirroring the uk population diversity data set and the point about um the interpretation of the data i know some some doctors and medics are kind of wary of at-home tests of all kinds whether that's you know fertility or genetic testing or a whole bunch of other things because you know a a doctor will interpret or a midwife or whoever will interpret information in a way that um with a lot more context than your average person who's got this at home and that sometimes maybe knowing too much can be as much of a problem as not knowing enough.

16:37How do you kind of manage that? Okay firstly the doctor typically has much less context than we get from our health assessment. Secondly while doctors tend to be Luddites when it comes to new technology. What we've done to counter that really is to just prove everything through clinical trials. We've done 29 research trials. The major hesitancy that clinicians see in primary care is that they're used to the gold standard of venipuncture, which is taking a lot of blood from the arm, which is problematic in a number of ways. Number one, a lot of people don't like big needles. Number two, 31 % of women are anemic.

17:22And so taking huge amounts of blood which is really unnecessary is actually quite harmful to them and number three physically getting to a clinic on the third day of your menstrual cycle means women end up putting it off so I think a lot of that is is just ignorance around where we've come in terms of capillary blood testing so we did multiple clinical concordance studies looking at the reliability of capillary blood versus venipuncture versus urine saliva dried blood spot menstrual blood we did them all um to show and prove and publish um the reliability of vitary i think it's an education piece and um actually i think now given the burden on systems many clinicians are leaning into remote diagnostic testing um we've never been one to say let's just test everything for this individual it's why our biomarker selection is dynamic and it's only based off symptoms you've told us so if someone tells us they have irregular cycles and have um of course she doesn't will test their androgens.

18:21We're not going to do that without a clinical reason to do so. So everything is tested with the view to clinical interpretation as opposed to overloading with lots of information. Got it. And when you say capillary, that's where the little kind of like farm finger prick, yeah, blood test rather than extracting a whole vial out of your arm at home, which would come with challenges, I imagine. What about, How, this is, you're obviously building Hedged City as a business. Why, do you think this should be part of our healthcare systems? Absolutely. I have had very recent calls. Actually, I've had multiple calls in order to try and get this integrated within national health systems.

19:10I think that's why for me, I thought if I can't get this in in the current journey that it is, I need to build internally as well so that if I can build this internally as well, that it's clinician facing, that it is removing a burden, removing the clinical workflow that is unnecessary, then I'll get buy-in. But in truth, trying to get it into the national health over the last few years hasn't been a problem. What we've built has been a problem with the system that is currently crumbling. I see with the new Women's Health Initiative, every single thing that they intend to do is something that we are doing, which is reducing time to diagnosis, reducing hospital burden, reducing unnecessary surgeries, or foreshortening the time to necessary surgeries for women.

19:57So I'm very hopeful for what's to come. Without saying too much. And you're building this business in the UK and Ireland. Would other places have been much easier countries to build it in? I'm wondering. I think we built it in the hardest jurisdiction there is, both in terms of the regulatory approvals here, but also in terms of healthcare attitudes. People regard their right to free healthcare like nothing else. They would rather die on a wait list than pay and be treated. And I really see that there is a reticence to pay for health care, which is why we've kept the price points so probably cripplingly low.

20:38But for me, that exercise in proving you can scale in a health care economy where health care is free means that we've overcome many of the burdens, many of the barriers that we we've fixed so many problems that you would face when you scale. So I'm very excited to take this globally with having built it in a system where people expect everything to be free. So, yeah, I'm excited for scaling it. How much does a test cost? Up until last year, it was£149 for the full health assessment, blood test and all of the clinically interpreted results from a gynecologist. So pretty cheap comparatively, more than three appointments in one.

21:22um we now include a um a call with one of our clinicians so it's 249 pounds um and I think that's that's helpful in terms of interpreting next steps for um for people and I guess that comes back to my earlier question around not just giving people a bunch of information that okay has some sort of annotated notes but there's a clinician who actually runs you runs you through it. Exactly. And I think that hand-holding, people want to see somebody, but the patient management system that we've built is very elegant. So it means that the clinician at least is armed with every single piece of information that has been told from the health assessment, the bloods, the ultrasound results, if they've had one.

22:05So yeah, it means that we can really help out the clinicians as well. What's been the hardest part of building this business? Florells just want to have funding. I mean, yeah, that's, I think that's probably every founder's problem, but really and truly trying to get funding for what is considered still niche is infuriating, actually. It doesn't matter how, you know, how you can rise to the ranks, the highest ranks of the highest universities in terms of academic, academic prowess. People still look at this as even falling within the term femtech. We don't, I do not identify as femtech. I identify as biotech, diagnostique.

22:51We have built something that is so much bigger than a test. It's a end to end clinical solution. It's building a data company. We're building a foundational model for women's health. I feel like we started off trying to be a baguette and we've twisted ourselves into a pretzel for every investor who's like you know it's not good to do d2c you should probably do b2b and we're like we're b2b now well you know you really need to you really need to get this like on a bigger level like a trade trying an insurance partnership we got an insurance partnership you know what's great is um software so like really you can scale software okay we've built software now and so i feel like we are we jump to every single thing to build and actually sometimes it can be hard to distill so much into to one presentation because we have there are multiple revenue streams that have underpinned our growth and I think that has protected us and then they'll come back and say you know you should really just focus on one thing so actually I kind of disregard some of the advice about focusing on one thing because it has stood to us inordinately to build so much across the reproductive life stages and has protected us in terms of having those multiple revenue streams especially in current economic climates what is the main revenue stream at the moment is it still the direct consumer or still direct consumer yeah and then you're when when will you introduce the software side of things as a revenue stream so we are we have a number of really exciting partnerships actually in the u.s and that is our next step so in the next 12 months the u.s and australia so in the next 12 months once we have our FDA clearance and approval we'll be able to then start to license the software as a medical device but we have some really exciting letters of intent and partnerships that really just help us to explode in terms of our growth and data and for me this is it's very exciting to be able to license our software as a medical device but the most exciting thing for me will always be returning to that clinical trial to say how can we improve the lives of women everywhere how can we reduce time to diagnosis and and actually what we've seen is that governments are screaming out for solutions that address all of those needs so i'm i'm hopeful for um timing to really help with uh just the urgency with which it's needed so in the u.s will that be you selling into hospital groups healthcare providers is that the customer yeah yeah and will they also offer fertility tests as part of that or yeah it's part of a tiered system so the first would be that health assessment that enables somebody to get to an appropriate panel of bloods to a diagnosis the second would be the logistics of us sending those blood tests the third tier is the full suite of assessment bloods interpretation and then the fourth is you know that full patient management system so we're aware the different end users whether your primary care or secondary care would require different levels of architecture in terms of that patient management system so we've made sure that the software as a medical device is quite modular to service all of those individual needs but the downstream saving for health systems for hospitals is quite unbelievable when you look at even one of the conditions that we screen for which is endometriosis, take on average nine years to be diagnosed.

26:22It costs the UK government 16 billion annually. And we can diagnose with 98 to 99 % confidence in eight days or screen for it. So if you take that that's just one condition affecting one in 10 and we can screen for 18, the saving to healthcare systems is inordinate. it and um i feel like expanding to the u.s is often this kind of holy grail of health tech um companies what has actually been you know just like a dream in comparison to working in the uk and ireland and what has been a lot harder i think with the u.s um attitudes are different energy levels are different you're met with enthusiasm with um so much more belief um i think there's a lot of reticence to get on board with new technology in the uk there's just there's number one there's there just isn't the funding here um but number two again you're always met with this comparison of the nhs which is you know a lot of people just feel a lot of maybe loyalty to despite the fact that it's not really serving a huge number of individuals so I am I just like the uh the enthusiasm of the U.S.

27:41and to embrace the things that are transformative and what about in the UK we've spoken a lot about what's been hard but has anything pleasantly surprised you about building the business here oh um yes I think our our users I think every single day I hear stories of people who say, I spent five years trying to get a diagnosis and I got one in a week with you. I'm now on treatment or I am now pregnant as a result of having done a fertility test. Or I met somebody the other day and she in waitrose and she said, both of my babies are fertility babies. So thank you. So those are the things that surprise me is not surprise me, but just keep me going.

28:22I think that will always be what makes me so resilient to any of the uphill battles. And every founder is facing an uphill battle. But those are the things that definitely keep me going every single day are just the end user stories because they are, you know, quite transformative. And I think with fertility and women's health, sometimes with, you know, people can say it's not life or death. But when it comes to family forming, it is life or no life. And it's nice to be able to help people have some what we call fertility hatchlings. And you mentioned this earlier about the challenges of raising for a sort of what was perceived as a sort of femtech or very women's business.

29:07what what what changes people's perspectives on that or is that not the way to go about it is it you have to find investors who see you as a health tech business as a biotech business as a you know something else is or or or have you been on a sort of journey with some investors where they you know you can help them overcome this perception that this is like a more niche thing than it is or a smaller you know tackling a smaller market or are those people you just you've learned to waste your time with don't waste my time with them we're very selective with our investors you either get it or you don't um and that's not something i think before we used to feel the need to educate around this but i think actually attitudes have really changed like we're now you know in a world where people are quite familiar with the statistics um we see a lot more openness um before i think i mean we see this time and time again you know they in vcs will arrange the call with their female intern or the only woman that they have on the team and the decision ultimately isn't made by the females in the team and so now we tend to look ahead and we'll be like right who's on the board i'm looking and seeing nothing but men so i'm not going to engage unless there is active interest here so i think that that sometimes is like making making your life a bit easier being like well ultimately who am i who am i pitching to is it an audience which is just actually quite sad you know it's 51 of the planet and all of those men have mothers sisters daughters aunts and wives so we shouldn't have to be so selective but unfortunately that's how we've learned to be is you know is there is there going to be a voice that actually understands this as a major opportunity and um i think i think what's boring for me is like you know everyone quotes the McKinsey report saying, you know, investing in women's health unlocks a trillion dollar opportunity.

31:04And actually that trillion dollar opportunity has severely been underestimated. It assumes, you know, this homogenous idea of just women's health as opposed to really breaking up the opportunity in menopause, perimenopause, bone health, cardiovascular health, in fertility, in pregnancy, in postpartum, in contraception, in all of those areas that are trillion dollar opportunities in isolation let alone combined so um yeah we're really trying to push beyond the rhetoric of it just being a mere trillion dollar opportunity it's it's actually a lot bigger what's the next when will you next need to raise funding and what what kind will that need to be always be raising um we are i think we i don't think we ever stop raising to be honest um but we are raising as we speak um so yeah that's really just going to unlock um honestly the opportunities that are already at our door that we just need money to service like the letters of intent we have for major corporations within the US so um we're actually quite excited about it um and the other aspect that we're excited about is really really when I look at some of the examples of these foundational models um that we've been ahead of the ball on that in terms of our data to build a foundational model for women's health so maybe that's an additional revenue stream that we can prop ourselves up with.

32:24But yeah, we never stop looking for investment. Exciting times ahead. And I think that's probably a good note to end on. So Helen, thank you so much for joining the show. You are welcome. And listeners, this will rather appropriately be my last podcast for a while as I am heading off on maternity leave. But never fear, I am leaving you in the capable hands of my colleagues, John Thornhill, Sifted's founder and the Financial Times' innovation editor, and Freya Prattie, who is an associate editor at Sifted. They will be holding the fort while I am changing nappies and mopping up. See you. As always, please rate, review and share.

33:04The podcast, this episode was produced by the wonderful Maya de Rumpel-Hornby.