What actually happens when a patient asks me about hormone therapy and breast cancer?

What happens is that she asks it sideways. Almost nobody says "what is my breast cancer risk on hormone therapy." She says "my sister told me it causes cancer," or "my mother was on it and then she got diagnosed," or she goes quiet after you say the word estrogen. The question is already in the room before you get to it, and if you wait for her to phrase it properly, she will leave with it unanswered and never fill the prescription.

The honest short answer, the one worth having ready in plain language: estrogen alone, in a woman who has had a hysterectomy, was not associated with an increase in breast cancer in the Women's Health Initiative and was associated with a reduction. Estrogen combined with a progestogen was associated with an increase, and the size of that increase in absolute terms was roughly eight additional diagnoses per 10,000 women per year of use in the WHI combined arm.

Eight in ten thousand is the number that does the work in the room, not the hazard ratio. This piece is about how to get to it without oversimplifying, and where the conversation tends to break down.

What the WHI arms found and why the two arms differed

The WHI ran two separate randomized hormone trials, and treating them as one study is the root of most of the confusion your patients arrived with.

The combined arm enrolled women with a uterus and randomized them to conjugated equine estrogens plus medroxyprogesterone acetate, or placebo. That arm was stopped early in 2002, and the breast cancer finding was the reason it made the front page. The reported hazard ratio for invasive breast cancer was approximately 1.24.

The estrogen alone arm enrolled women who had had a hysterectomy and randomized them to conjugated equine estrogens or placebo. It ran longer and reported a hazard ratio for breast cancer below 1, in the region of 0.77 to 0.79, meaning fewer diagnoses in the treated group.

Two arms, two directions. The most common reading of the difference is that the progestogen component, not the estrogen, drove the breast signal in the combined arm. That reading is supported but not proven, and the WHI used one specific progestogen, medroxyprogesterone acetate, at one dose. Whether micronized progesterone carries the same signal is an open question rather than a settled one. Observational data has suggested it may be more favorable. Randomized data at the same scale does not exist and probably never will.

The 2019 collaborative reanalysis

Be ready for this one, because it produced a second wave of headlines. The Collaborative Group on Hormonal Factors in Breast Cancer published a large meta analysis of epidemiological evidence in 2019 reporting duration dependent excess risk with both estrogen alone and combined regimens, which is a less reassuring picture than the WHI estrogen only arm.

You do not need to resolve that tension for a patient. You need to name it. The randomized trial and the pooled observational evidence disagree about estrogen alone, they agree that combined therapy carries an increase, and both find the absolute magnitude small in any single year.

Keep reading: How do I actually price a 60 minute menopause consult in a cash pay clinic?

Absolute versus relative risk in numbers a patient can hold onto

A 24 percent increase sounds catastrophic. Eight in ten thousand does not. Both describe the same finding. Your job in that moment is arithmetic, out loud, on paper.

Take the WHI combined arm figure of roughly eight additional invasive breast cancers per 10,000 women per year of use. Convert it.

• Eight per 10,000 women per year is one additional diagnosis for every 1,250 women treated for one year.
• Treat 1,000 women for five years and, holding the yearly rate flat, you would expect about 40 additional diagnoses across that group, or roughly one woman in 25 of that thousand.
• Framed for the individual: an added chance of roughly 0.08 percent per year, or about 0.4 percent over five years.

State the assumptions out loud, because they are assumptions. That arithmetic holds the annual rate constant across five years, which the WHI data suggests is not quite right for the combined arm, where the separation between curves widened with duration rather than staying flat. It also applies a trial average to a specific woman whose baseline risk may be well above or below it. Say both of those things. Patients trust arithmetic more when you show them where it bends.

The comparison that helps most is to her own baseline. Lifetime breast cancer risk for a woman in the United States is commonly cited as roughly one in eight, and annual risk rises with age regardless of what you prescribe. This is a small addition to a number that was never zero.

Estrogen alone after hysterectomy versus estrogen plus progestogen

The single most useful sorting question in this conversation is whether she has a uterus.

If she does not, the breast conversation is materially different and often more favorable than she expects. Many women who have had a hysterectomy have spent years believing they carry the combined arm's risk, because that is the study that got reported.

If she does, she needs a progestogen for endometrial protection, and that is not optional. Unopposed systemic estrogen in a woman with a uterus produces endometrial hyperplasia and carcinoma at rates that dwarf the breast discussion you are having. The progestogen is the thing that makes the estrogen safe somewhere else.

Situation	Regimen	WHI breast signal	What she usually assumes
Prior hysterectomy	Systemic estrogen alone	Hazard ratio below 1 in the estrogen alone arm	That the 2002 headline applies to her
Uterus intact	Estrogen plus a progestogen	Hazard ratio around 1.24 in the combined arm	That the progestogen is optional
Genitourinary symptoms only	Low dose vaginal estrogen	Not the population studied in either systemic arm	That any estrogen carries the same risk

The third row deserves emphasis. Low dose vaginal estrogen is a local therapy with minimal systemic absorption, and it is not the intervention either WHI arm tested. Patients, and a fair number of referring clinicians, apply the systemic findings to it anyway. The boxed warning on the packaging does not help you here, and it is worth pre empting the moment she reads the insert at home.

Keep reading: What does the FDA boxed warning on estrogen actually require me to tell patients?

Duration, timing of initiation, and the window hypothesis

Two variables change the calculation more than anything else on the prescription: how old she is when she starts, and how long she stays on.

The timing hypothesis holds that initiating within about ten years of the final menstrual period, or before age 60, produces a more favorable overall benefit to risk profile than initiating later. The Menopause Society's position statement reflects this framing. It is derived largely from subgroup and secondary analyses rather than from a trial designed to test it, which is a real limitation and one you can state without undermining the point.

Duration is the more concrete lever. In the combined arm the breast signal emerged with continued use rather than immediately, which is why the practical framing is not "should I take this forever" but "let us plan to revisit this every year and decide again."

That reframing does more for adherence than any statistic. A patient who believes she is committing to a fifteen year decision will often decline. A patient who understands she is committing to a twelve month trial with a scheduled review will usually start.

Screening intervals, breast density, and what changes on therapy

Practical points that come up after she says yes.

Combined therapy can increase mammographic density, which raises the rate of recalls and additional imaging. That means more phone calls about a callback and more anxiety, not necessarily more cancer. Tell her before the first mammogram on therapy, not after the callback letter.

Screening intervals do not change simply because a patient is on hormone therapy. Follow the guideline your practice follows for her age and risk profile. If she has dense breasts and your state has a density notification law, she has probably already received a letter, and the supplemental imaging conversation is separate from the hormone one even though she will merge them.

See how PauseNotes handles this for menopause and midlife women's health clinics

Where the conversation usually goes sideways

Four recurring failure points.

Leading with the hazard ratio. The relative number lands as fear and cannot be un heard. Give the absolute number first, then the relative one if she asks.

Dismissing her family history too fast. A patient whose mother had breast cancer is not asking a statistical question. Take the history properly, and if it warrants a formal risk assessment or genetics referral, that changes the conversation entirely rather than shading it.

Overselling bioidentical framing. If you imply that a compounded or body identical preparation removes the breast question, you have made a claim the evidence does not support and you have taught her that this discussion is negotiable.

Ending it in one visit. She will go home, read something, and change her mind. Build the follow up in.

Documenting the discussion and the shared decision

Your note should show a decision, not a recitation. Six elements cover it.

1. Her stated concern in her own words, including who told her what.
2. Uterus present or absent, and the regimen that follows from it.
3. The absolute risk figure you gave her and the source you attributed it to.
4. Baseline breast history, family history, and date of last mammogram.
5. The planned duration and the date of the scheduled reassessment.
6. Her decision, including a decision to defer.

That last item matters more than it looks. A documented "declined for now, will reconsider at six months" is a clinical position. A blank is a gap that reads badly later.

What makes the reassessment work

Every part of this conversation rests on a promise: that you and she will look at whether it is working and decide again on a schedule. That promise fails quietly when the twelve month visit arrives and neither of you can reconstruct what the last year actually looked like.

PauseNotes sends structured symptom scores to your patients between visits and charts them, so the reassessment starts from a line on a page rather than from "how have you been doing." When the risk you counseled on is small and the benefit is what justifies it, being able to show her the benefit is the whole argument.