Women’s health is having a moment. Lots of moments. And this moment is a test case for many stakeholders. Will investors move dollars into this space? Will the definitions of “success” be amended to account for advancements that benefit all of us or just some of us? Will anybody change the definition of “atypical” chest pain to simply refer to the most common presentation of myocardial infarction in 51% of us?
How women present with chest pain should probably be labeled as “just typical.”
When it comes to healthcare consulting, a different kind of test is underway. Digital transformation, the integration of artificial intelligence, smoother regulatory pathways: all great. If they are built upon complete, usable, and actionable evidence. Transformation won’t be truly transformational unless the right building blocks are used from the start.
For those of you who missed the headlines, women were excluded from clinical trials in the US in the late 1970s due to a confluence of factors. The natural fluctuation of hormones was considered a confounding variable that complicated research. The thalidomide tragedy shone a spotlight on the risk to an unborn fetus. Politically, socially and culturally there was simply no pushback to a paternalistic concern about protecting half the population. Underlying this was a deep misunderstanding about the importance of sex-based research when it came to healthcare. It turns out, we are not all 70-kg white men. Sex-differences affect how we respond to disease, how we respond to treatment. Of the FDA-approved drugs that have been pulled from the market since 1997,guess how many were pulled due to real-world evidence of adverse effects in women? Eighty percent.[1] Because they were never studied in a population that might react differently.
This lack of interest in conditions that affect women uniquely, differently or disproportionately left a gaping hole. Nature abhors a vacuum. It must be filled. And so, it has been. With “innovations” that may or may not be beneficial. May or may not be harmful. May or may not be rooted in evidence.
The flood of digital solutions for women’s health ranges from the utterly unsubstantiated to the rigorously built and clinically validated. For example, take the claims of real-time noninvasive hormone monitoring. For the average lay person, it is difficult to understand why we can monitor glucose continuously but the same technology doesn’t work for hormones. (It doesn’t, by the way.) Pulling one step further away from CGM, “noninvasive” hormone monitoring sounds great but that nirvana has not yet been reached and validated. Despite the devices that claim to offer such insights. And that is an issue.
Test problem number one: of the wealth of “solutions” for women’s health, which ones are going to move the needle for real?
Now suppose we generate solutions that are rooted in science. Fabulous. But let’s ask an important question.
Whose science?
The data is incomplete. It wasn’t until 1993 that all NIH funded research was required to include women.[2]Even then, there has been a lack of research into sex-differences across the entirety of human health conditions. Sex-differences in research were considered secondary variables and their importance largely discounted.
Think about that. The decades of information upon which we are basing innovation is not decades of research inclusive of women. This is a very important point. When data is incomplete:
- Risk is underestimated
- Benefit is uncertain
- Variability in patient outcomes is minimized
Test problem number two: we aren’t starting from zero. But we are starting with inherently biased baseline data. Arguably more dangerous.
Not that there isn’t any research specific to women. But now we enter another area that is problematic. What evidence we have is fragmented. Women’s health is inherently longitudinal. Yet science has historically treated it as a series of isolated events. That is an issue. When you disaggregate data across time, settings and specialties, you miss a trick.
For example,
- Cardiovascular risk models that are based on male-dominated data sets
- Reproductive health as a stand-alone condition, not linked to long term health outcomes
However, we know that hypertensive conditions of pregnancy affect long-term cardiovascular risk.[3] We know that menopause represents a rapid increase in cardiovascular risk for women.[4]
In order to incorporate a reproductive history with long-term cardiovascular risk, all that data needs to be integrated. How is that possible if reproductive health is sitting in a different bucket from cardiovascular health data that was largely generated on men?
Test problem number three: fragmentation doesn’t just slow down innovation, it fundamentally limits what we know from the start.
And the list goes on. How can we scale innovation when the assumptions at the beginning are incorrect? Models trained on partial data grow more erroneous at the same time they grow more confident. In a world of growing technological integration in healthcare, data drift is a thing. If the underlying data is already shaky, we’re talking rapids, not drifts. Innovate on partial truths? Don’t act surprised when your outcomes are inconsistent.
If the test for healthcare consulting is whether we can build meaningful innovation on incomplete, fragmented, and biased data, then the answer is straightforward: we can’t. Not reliably.
So what do we do with this? If the problem is structural, the solution cannot be another layer of tools. It has to be a rebuild of the foundation. For healthcare consultants, this creates a non-trivial risk: optimizing systems built on incomplete data doesn’t improve care. It scales error.
This requires three non-negotiable shifts.
- Sex-specific data as a non-negotiable
- Longitudinal, lifecycle integration of that data
- Data embedded into solutions. Validated solutions.
There is a reason in medicine as to why young doctors are trained to take a thorough medical history. The information we glean is relevant to the decisions we make going forward.
Advancing healthcare technology is no different. The past informs the future. If we don’t start with solid, evidence-based building blocks, the outcome is worthless.
[1] FDA. 2001. “Drug Safety: Most Drugs Withdrawn in Recent Years Had Greater Health Risks for Women | U.S.” GAO.
[2] Wood, Susan. 2024. “History of Women’s Participation in Clinical Research.” Office of Research on Women’s Health.
[3] Cederlof, Elin T., Maria Lundgren, Bertil Lindahl, and Christina Christersson. 2022. “Pregnancy Complications and Risk of Cardiovascular Disease Later in Life: A Nationwide Cohort Study.” Journal of the American Heart Association 11, no. 2 (January).
[4] El Khoudary, Samar R., Brooke Aggarwal, Theresa M. Beckie, Howard N. Hodis, Amber E. Johnson, Robert D. Langer, Marian C. Limacher, JoAnn E. Manson, Maria L. Stefanik, and Matthew A. Allison. 2020. “Menopause Transition and Cardiovascular Disease Risk: Implications for Timing of Early Prevention: A Scientific Statement From the American Heart Association.” Circulation 142, no. 25 (November).