Why endometriosis still takes years to diagnose

Endometriosis affects roughly one in ten women and those assigned female at birth of reproductive age. Endometriosis UK reported an average UK diagnosis time of 8 years and 10 months in 2024, up from 8 years in 2020. [1] [2]

That delay is not a curiosity. It changes who is available to study, what the evidence base contains, and how findings should be read.

Key points

• Long diagnosis delays change who becomes visible to researchers.
• Samples based only on confirmed diagnosis can exclude years of lived experience.
• Recruitment design should account for suspected and undiagnosed cases where appropriate.

Why the delay persists

Pain is normalised early. Girls are told that severe period pain is part of being a woman. Many wait years before raising it at all, because they believe there is nothing to raise. This connects to a wider pattern described in the gender pain gap evidence.

Symptoms are non specific and cross specialties. Pelvic pain, bowel symptoms, bladder symptoms, fatigue and subfertility do not point to one department. Women are moved between gynaecology, gastroenterology and general practice, and each handover resets the clock.

Hormonal treatment masks the picture. The combined pill is often prescribed for painful periods without investigation. It can help symptoms, which is good for the patient in the short term and unhelpful for diagnosis, because improvement gets read as resolution.

Imaging misses it. A normal scan does not rule out endometriosis. Superficial disease is frequently invisible on ultrasound and MRI, and a normal result is too often communicated as an all clear.

Laparoscopy is a surgical decision. For a long time the definitive answer required surgery, and surgery is rationed, delayed and, understandably, not offered lightly to young women.

What the delay does to the evidence base

Recruit for an endometriosis study by requiring a confirmed surgical diagnosis and you have selected for a specific population: women who persisted, who were believed, who reached a specialist, and who were offered surgery. That group is older at the point of study, more likely to be white, and more likely to be in a position to advocate for themselves in a clinical setting.

Everyone still in the diagnostic gap, which on these numbers is a large share of everyone with the disease, is absent. Research designed this way describes the diagnosed population and then gets cited as though it describes the disease.

This is the same structural problem described in the women's health data gap: the missing data is not random, and the people missing from it are the people the system already failed.

What better study design looks like

• Include suspected as well as confirmed endometriosis, and record which is which so the analysis can separate them.
• Capture time from first symptom and time from first medical contact, not just time since diagnosis. The two delays have different causes and different fixes.
• Record every specialty seen along the way. Pathway data is where the intervention points show up.
• Recruit outside clinic lists. Anyone who disengaged from services will not be on one.
• Report the diagnostic status of your sample in the abstract, not buried in the methods.

Why it matters commercially too

For anyone building products, services or apps in this space, the diagnosed population is not the market. The much larger group is women who know something is wrong and have no label for it. Research that only speaks to diagnosed women will systematically misread demand, language and unmet need.

If you are designing a study in this area, our guide to recruiting for women's health research covers how to write criteria that do not accidentally exclude the undiagnosed.

References

1. Endometriosis UK. Endometriosis facts and figures.
2. Endometriosis UK. Endometriosis diagnosis times still unacceptably long and getting longer. 2024.