Most Women See Three Doctors Before Getting An Answer

Xella Health CEO Kelly Lacob on the gap between finding a condition and treating it.

Researchers at the University of Copenhagen pulled 21 years of hospital records covering 6.9 million people, the entire Danish population, and compared the age at which men and women received the same diagnoses. Across more than 700 conditions, women were diagnosed later. The average gap was about four years.

The 2019 Nature Communications study, led by Westergaard and Brunak, is observational; its authors noted men tend to present to care later in general, which would widen the true gap further.

It is the finding Xella Health cites, and it describes a detection gap.

The measurement problem has a solution now. The system it lands in does not.

Xella Health launched on June 24, 2026, out of San Francisco. It closed $3.7 million in pre-seed capital in January led by Precursor Ventures, with Capital F, Ulu Ventures and Swizzle Ventures participating, and puts total angel and pre-seed funding at $4.7 million.

Kelly Lacob co-founded it with Adriana Dantas and Jesus Ching, both of whom she worked alongside at Mammoth Biosciences, the CRISPR diagnostics company. Dantas came through Roche and Becton Dickinson on commercial operations. Ching runs technology. Lacob had stepped back from Mammoth to move her mother into hospice with late-stage ovarian cancer when Ching brought the idea to her and Dantas.

The product runs a venous draw against more than 100 biomarkers spanning hormones, proteins, metabolic and nutrient markers, layered with questionnaire data on symptoms, history and risk. The company says it screens for more than 130 conditions including endometriosis, PMOS, Hashimoto's, PMDD and perimenopause staging. Panel turnaround is two to three weeks. A clinician reviews every report before it reaches the member.

During beta, several reports went out before the telehealth layer was operational. Some carried serious findings. Those members took the reports to their own physicians.

"It was just so demoralizing to hear from them that they got turned away from their GP or from others," Lacob told The Luteal. She recounted what members were told. "I don't really know what to do with this information."

"We do know what to do with it," Lacob said. "And yes, it is actionable."

The physicians in that account were being asked to act on conditions most clinicians were never trained to manage.

Kling and colleagues surveyed residents in family medicine, internal medicine and obstetrics and gynecology across 20 US residency programs for Mayo Clinic Proceedings in 2019. Of the 177 who answered the question, 12 said they felt adequately prepared to manage women going through menopause. That is 6.8 percent. One in five had received no menopause lecture during residency at all. The response rate was 26 percent, which skews the sample toward trainees engaged enough to reply.

The pattern holds across the conditions Xella screens for. Gibson-Helm, Teede, Dunaif and Dokras surveyed 1,385 women with PCOS across 32 countries for the Journal of Clinical Endocrinology & Metabolism in 2017.

Nearly half saw three or more health professionals before anyone named the condition. For a third it took more than two years. Once diagnosed, 15.6 percent were satisfied with the information they were given about it. Respondents were recruited through patient support groups and diagnoses were self-reported, so the sample skews toward difficult experiences.

There is a second reason, legal and structural barriers.

A clinician who did not order a test has, in most circumstances, has no duty to act on it.

Arizona's direct-access testing statute says so directly, and also bars the laboratory from billing a third-party payor for a test run without a provider's order. State law varies, and malpractice carriers advise that duty can attach where a provider-patient relationship already exists. Federal routing guidance points the same way, advising that results need not go to a patient's primary care physician unless that physician ordered the test or agreed to handle follow-up.

A woman arriving with a 100-biomarker report is asking a physician with no ordering relationship, no chart context, no billing pathway and no clear duty to assume responsibility for someone else's output.

How often this happens is not well measured.

The closest available read is a 2019 survey of 1,502 Kaiser Permanente physicians published in the Journal of Personalized Medicine, which drew a 15 percent response rate and covered genetic results rather than biomarker panels. Among the 35 percent who had received a consumer test result from a patient in the prior year, 40 percent made at least one referral, and 78 percent of those referrals went to clinical genetics.

A contracting workforce only worsens this friction. HRSA's Health Workforce Simulation Model projects that the OB-GYN workforce will meet only 81.7 percent of national demand by 2037, with a projected 46 percent shortage in rural areas by 2038. An overburdened specialty cannot easily absorb an influx of unrequested, unbillable lab results.

Recognizing that gap, Xella built the clinical layer itself.

Every member is assigned a dedicated clinician, mostly OB-GYNs trained in women's health, who reviews the findings and the proposed plan and can approve, edit or reject it.

The clinician network is both the largest variable cost and the constraint on how fast the consumer product can scale, which makes the relevant unit economics clinical cost per member and clinician capacity per member rather than tests processed.

Lacob points out that standard lab panels carry inherent limitations. Population averages fail to account for personal baselines, life stage, or sex-specific differences, and individual markers are rarely evaluated for what they contribute to an overall pattern.

"Personal baselines are very important, because a lot of these things, our baseline is different," she said. "And so knowing when those small changes are happening."

She frames the old constraint as human rather than technical.

"A lot of the limitations were based on how much knowledge can one individual hold in their brain at a time and how many different connections can they make, and to think about a specific organ system or specific subset of their specialty focus," she said. "But when you have AI at your disposal, you can look at orders of magnitude more data and make sense of it within a matter of moments."

That perspective offers an evolved approach to a medical screening model historically built around single markers. Established guidelines, such as those from the US Preventive Services Task Force on thyroid and vitamin D screening, have typically evaluated tests individually to weigh benefits against unnecessary treatments.

Leaders like Lacob see an opportunity to expand on that foundation by looking at biomarkers as interconnected patterns rather than isolated data points.

The key to scaling proactive screening lies in effectively managing follow-up care. A 2019 JAMA Network Open study found that 99 percent of internists have experienced a "cascade of care" triggered by incidental test findings, with 94 percent reporting instances where those follow-ups yielded no clinical benefit.

Addressing those provider concerns is a crucial step toward building broader clinical adoption. With trust.

The answers exist now.

But the surrounding medical and regulatory infrastructure is just not there yet.