Case 003 — sleep, refused hormones, nine weeks in.

Encounter summary

Member is a 51-year-old administrative director at a non-profit. Twelve months of progressively-worsening sleep — initial-onset insomnia, multiple night wakings (3–6 per night), 03:14 a.m. wake-up that has become predictable enough to be a private joke. Vasomotor flushes 4–7 per day plus night sweats heavy enough to require a change of pajamas. PHQ-9 score of 11 (moderate); GAD-7 of 8 (mild). Cycle: skipped two cycles of the prior six. No depression history; no anxiety history; no prior psychotropic use.

Initial visit framing was straightforward. The Menopause Society 2022 statement¹ supports menopausal hormone therapy in this exact profile (under 60, within 10 years of probable FMP, symptomatic, no contraindications). Armitage made the case for transdermal estradiol 0.05 mg/day plus oral progesterone 200 mg at bedtime as first-line. Member declined hormone therapy — partly informed by a family history of breast cancer in a maternal aunt diagnosed at 67, partly by personal preference. The decision was not negotiated; it was the member's. The shared plan that emerged: CBT-I for insomnia plus venlafaxine 75 mg for the vasomotor symptoms (an off-label but well-evidenced use²), and a re-conversation about MHT at week 12 if the symptoms remained intolerable.

Splits

Split	Week	Event	PHQ-9	Sleep (h)	VMS/d
00:00	W00	Initial visit, MHT declined; venlafaxine 37.5 mg started	11	4.2	5–7
02:00	W02	Venlafaxine 75 mg; CBT-I session 1/8 with Wahlquist	10	4.4	4–6
04:00	W04	Sleep window restriction (23:00–05:30); CBT-I 3/8	9	4.8	4
06:00	W06	Stimulus control consolidating; CBT-I 5/8	8	5.4	3
08:00	W08	Sleep window expansion to 22:30–05:30; CBT-I 7/8	8	5.7	3
09:00	W09	Co-visit (Armitage + Nair); MHT re-conversation requested	8	5.6	3

Debrief — what was tracking

Two things tracked partway. The CBT-I work — running through Dr. Idris Wahlquist on Lake Avenue, who has run a structured CBT-I program for over a decade — moved both the sleep latency and the night-wakings count meaningfully. By week eight the member was sleeping in a four-and-a-half-hour consolidated block, which is an improvement on the four-fragments-of-an-hour they began with. The venlafaxine titration worked too: vasomotor symptoms dropped from 5–7/day to a steady 3/day by week six, and the night sweats dropped from drenching to mild. The PHQ-9 trajectory bent in the right direction.

What didn't track

The 03:14 wake-up did not yield. It is still happening four nights of seven. The member's total sleep time is still under six hours. The PHQ-9 remains in the moderate range. They are tired in the way that is not solved by time. At week nine the member opened the MHT conversation again, and we are scheduling a longer follow-up to walk through the family-history calculus with them, against the WHI long-term mortality data³, the IBIS-II family-history adjustments, and the option of fezolinetant as a non-hormonal alternative.

Plan, T+0

• One-hour follow-up booked W11; family-history calculus is the agenda.
• If MHT initiated: transdermal estradiol 0.05 mg/day patch + oral progesterone 200 mg HS; baseline labs already in chart; follow-up W14.
• If fezolinetant initiated: 45 mg daily; LFTs at baseline, four weeks, six months per FDA label.
• If neither: continue venlafaxine, continue CBT-I maintenance with Wahlquist, re-evaluate at W18.
• Flag remains amber until member feels rested. We are not leaving it here.

What this case is and isn't

This is the case we wanted on the splits board because it is amber. The flag is not a failure flag — the venlafaxine and the CBT-I have done real work, and the member's mood has improved meaningfully — but it is not a green flag, because the sleep is not fixed. The temptation in the perimenopause lane is to declare a partial win and move on. We don't. The case logs the flag honestly and the next visit is on the calendar.