—

On personalising healthcare with AI

March was International Women's Month, and this year, we targeted women who'd experienced healthcare bias, been dismissed at the hospital or misdiagnosed. Women who've been called crazy, or told their symptoms were just in their heads.

This included my mum, who's story was the opening one that set the tone for the campaign, which further explains why it remains a dear project to me. Based on the stories shared, below is a categorised summary of the most common reasons for the delays in diagnosis, along with examples of how they showed up:

Category Definition Example(s)
Dismissive Attitude Healthcare providers downplayed or ignored symptoms. “Doctors told my mom I was probably pregnant — turned out to be a brain tumor.”
“You’re young and healthy.”
Misdiagnosis Incorrect diagnosis that delayed correct care. “Misdiagnosed with IBS, turned out to be a parasitic infection.”
“Told it was a pulled muscle, but my lungs had collapsed.”
Lack of Adequate Testing Delays in running proper diagnostics or reliance on limited tests. “Doctors didn’t run tests. Ultrasound found ovarian cyst causing internal bleeding.”
“Finally got diagnosis in one test that took 2 hours.”
Gender or Age Bias Pain and symptoms dismissed due to gender or age stereotypes. “You're too young to be in this much pain.”
“Women can’t have serious issues at 23.”
Psychologising Symptoms Symptoms attributed to anxiety or depression without ruling out medical causes. “Told it was stress, not PCOS.”
“Given antidepressants, turned out to be encephalitis.”
Late or No Specialist Referral Delay in referring to a qualified specialist. “Saw 14 doctors before one referred me to a specialist.”
“Needed an orthopaedic surgeon, was sent to neurologist and geneticist instead.”
Over-Reliance on Stereotypes Doctors relied on generalisations (e.g., pregnancy cures pain). “Was told birth control would solve my issues — later diagnosed with endometriosis.”
Inadequate Communication Patients were not clearly informed or educated about their condition. “No one explained the cause of pain. Just sent me away with meds.”
Stigma or Moral Judgement Patients judged for symptoms linked to sexual or reproductive health. “Called a prostitute at 16 for asking to see a doctor about suspected infection.”

These not only delayed diagnosis but also caused emotional distress and long-term health consequences. While the human factor in healthcare will always be of value, the pattern seems to confirm something many in healthcare — especially the old heads — aren't comfortable with; the over-reliance on human interaction isn't healthy for both doctors and patients.

Starting with Delayed Diagnoses and Dismissals

The struggles women face in getting timely and accurate diagnoses aren’t just anecdotes — they’re symptoms of systemic issues baked into healthcare. One major factor is the historical bias in medical research, which has long prioritised male bodies as the default.

Conditions like polycystic ovary syndrome (PCOS), endometriosis, and adenomyosis, which disproportionately or exclusively affect women, have been understudied relative to their prevalence and impact. This knowledge gap trickles down to clinicians, who may lack the training or awareness to recognise these conditions early, if at all.

Here are 20 of the 43 conditions we've recorded from women within a week of our #notinyourhead project.

1. Polycystic Ovary Syndrome (PCOS) — A hormonal disorder causing enlarged ovaries with small cysts, leading to irregular periods, excess androgen levels & insulin resistance. https://t.co/8igu6KX72F

— Famasi Africa (@FamasiAfrica) March 16, 2025

Then there’s the dismissal factor. Women’s symptoms are too often chalked up to “hormones,” “stress,” or even “hysteria” (a term that’s thankfully fading but whose echoes linger). Studies have shown that women are more likely than men to have their pain minimised or misattributed to psychological causes.

  1. 2000 Study on Heart Attack Misdiagnosis (Pope et al.)
    Published in the New England Journal of Medicine, this study found that women were seven times more likely than men to be misdiagnosed and discharged during a heart attack. Chest pain in women was more often attributed to anxiety or psychological distress, delaying critical treatment. This reflects a broader tendency to psychologise women’s pain, even in life-threatening situations.

  2. 2001 Study on Gender Bias in Pain Management (Hoffmann & Tarzian)
    Published in the Journal of Law, Medicine & Ethics, this study reviewed existing literature and found that women’s pain reports were more frequently dismissed as emotional or psychological rather than physical. It highlighted cases where women with chronic pain conditions, such as fibromyalgia, were prescribed sedatives instead of pain medication, while men with similar complaints were more likely to receive analgesics. The authors pointed to a pattern where healthcare providers perceived women’s pain as less credible, often attributing it to stress or anxiety.

  3. 2008 Emergency Department Study on Acute Abdominal Pain (Chen et al.)
    In a study published in Academic Emergency Medicine, researchers analysed 981 emergency room visits for acute abdominal pain. They found that women were up to 25% less likely than men to receive opioid painkillers, despite reporting similar pain levels. Women waited an average of 65 minutes for analgesics, compared to 49 minutes for men. The study suggested that providers might attribute women’s abdominal pain to gynaecological or psychological issues rather than treating it as urgently as they would for men.

  4. 2016 Study on Chronic Pain Judgments (Schäfer et al.)
    Published in Pain, this research examined how healthcare providers judged chronic pain patients. It found that providers were more likely to believe female patients exaggerated their pain compared to male patients, even when pain levels were comparable. Providers often recommended psychotherapy for women but prescribed opioids for men, implying a bias toward viewing women’s pain as psychological rather than somatic.

  5. 2021 Study on Gender Biases in Pain Estimation (Zhang et al.)
    Published in the Journal of Pain, this study involved observers rating pain in male and female patients based on video clips of real patient experiences. When patients expressed identical pain levels, observers rated women’s pain as less intense and were more likely to suggest psychotherapy over medication, while men’s pain was taken more seriously and linked to physical causes. The study used objective measures like facial expression analysis to confirm the bias wasn’t due to differences in pain expression.

Collectively, these illustrate a recurring pattern: women’s pain is often underestimated or reframed as psychological, leading to disparities in treatment compared to men. The biases appear rooted in stereotypes about women being more “emotional” or prone to exaggeration, rather than in objective differences in pain presentation.

One of the women with chronic pelvic pain was told it was “just bad periods” for years before a doctor considered endometriosis, which affects roughly 1 in 10 women and can take 7-10 years to diagnose on average. This isn’t just frustrating — it’s debilitating. Untreated, these conditions can lead to infertility, chronic pain, or worse.

The emotional toll compounds the physical suffering. Imagine visiting doctor after doctor, each one shrugging off your symptoms, leaving you questioning your own body. The young mother with adenomyosis who faced seven dismissals didn’t just lose time — she lost trust in the system meant to help her. And when diagnoses do come, they’re often late, after irreversible damage has set in.

For PCOS, delayed recognition can mean missed opportunities to manage metabolic risks like diabetes. For endometriosis, it can mean scar tissue buildup that surgery can’t fully undo. The stakes are high, and the status quo isn’t cutting it.

Sadly, expecting more from clinicians is like finding a country no Nigerian has been to — very possible, but betting on it isn't profitable. Not due to lack of care in many cases, as there are lots of clinicians who're good at what they do and aren't clouded by bias, but due to being overwhelmed and under-equipped.

Well, it's the 20th century and Artificial Intelligence is a thing now. No, this isn't to say AI doesn't have its bias, or that it's pitch perfect. It's still being developed. But even at its current capabilities, it's very less likely to call women prostitutes for seeking medical help on sexual health. And if we're intentional about fine-tuning it enough, the rate at which women get misdiagnosed with anxiety when it's really PCOS would reduce.

So yes, AI-Powered Healthcare

I get it, even Harry Potter's magic wand didn't work at times. AI isn’t a magic wand, but it’s a damn good tool to tackle these problems. The issue is that many interact with it today like it's a replacement for humans when in reality, it's a thought partner; something that enhances. For example:

Tying It Together

The problem isn’t just delayed diagnoses — it’s a mix of missed opportunities, eroded trust, and preventable suffering rooted in outdated systems and biases. AI offers a shot at rewriting that story. It’s not about replacing doctors/clinicians, but equipping them with precision tools — better imaging, smarter symptom analysis, custom treatments, and even a mirror to reflect their own blind spots. If we get this right, the woman with PCOS doesn’t wait years for answers, the endometriosis patient skips the dismissal merry-go-round, and the adenomyosis mom gets help before her seventh doctor. That’s not just healthcare improvement — it's an experience worth rooting for.