Nine Years of Eszopiclone, and a Patient Who Won’t Try Anything Else
Nine years on the same drug, no dose escalation, no adverse events — and a flat refusal to try the therapy every guideline says should have come first. The disagreement is about what “working” actually obligates anyone to do next.
B.F., a 67-year-old woman, retired from teaching high school English four years ago after a thirty-four-year career, and now spends much of her time in a book club she helped found and visiting her three grandchildren, who live nearby. She has been on eszopiclone 2 mg nightly for insomnia since she was 58, prescribed originally by a physician who has since retired from the practice, at a point in her career when grading loads and a difficult department transition had made her sleep genuinely unmanageable. She describes that period, looking back, as the only time in her adult life she felt truly desperate about not sleeping.
She has no psychiatric history beyond the original insomnia diagnosis, and her only other medical condition is well-controlled osteoarthritis in both hands, managed with occasional acetaminophen and no other regular medications. In the nine years since starting eszopiclone, she has had no falls, no cognitive complaints from herself, her husband, or her physician, and no dose increases of any kind — the same 2 mg, nightly, without interruption. She sees her current physician twice a year for routine care and describes her sleep, unprompted and with evident relief, as “the one thing that still works exactly the way it should,” a contrast she draws deliberately against the arthritis she has had to keep adjusting to.
When CBT-I was raised at her last visit — part of a practice-wide push to periodically review long-term hypnotic use — she declined immediately and without much discussion, saying plainly that she has “no interest in relearning how to sleep” after something has worked reliably for nearly a decade, and that she would rather continue the medication exactly as it is than commit to a structured behavioral program she didn’t ask for and, in her own words, doesn’t believe she needs.
Twice-yearly follow-up, nine years in
The AASM’s own guidance on long-term hypnotic use calls for periodic reassessment, not indefinite, unexamined continuation — and nine years without ever attempting the treatment named as first-line in every relevant guideline is a real gap, even in a patient who is, by every visible measure, doing well.
I’d want at least one honest conversation about why CBT-I specifically feels unwelcome to her, separate from just re-offering the same referral she’s already declined.
I hear the guideline argument, but I think it’s being applied here as though the goal is procedural compliance rather than patient outcome. She is 67, has taken a low, stable dose for nine years with zero adverse events, reports excellent functional sleep, and has already declined the alternative clearly and without ambivalence.
The AASM guidance you’re citing is about periodic reassessment of ongoing need and safety — which we’re doing, right now, in this visit — not a mandate to keep re-offering a specific therapy a competent adult has already declined. Continuing to press CBT-I after a clear refusal risks reading as not respecting her stated preference more than it reads as good medicine.
I don’t think this needs to resolve into “keep offering” versus “drop it entirely.” Long-term eszopiclone at a stable, low dose in an older adult with no adverse events is a genuinely reasonable thing to continue — the pharmacologic argument for stopping a drug that’s working, in someone who has tolerated it this long without escalation, is weaker than it would be for a higher-risk agent.
Where I’d land is: document her refusal clearly, continue the medication, and revisit the offer — once, briefly, without pressure — only if something actually changes, a fall, a cognitive concern, a new diagnosis that shifts the risk-benefit picture. Repeating an unwanted referral at every visit isn’t reassessment, it’s just asking the same question until she gives a different answer.
Agreed: continue eszopiclone at the current dose, document her clear and informed refusal of CBT-I, and avoid re-offering it as a routine matter at every visit.
Not agreed, and named explicitly rather than smoothed over:
A guideline-recommended first-line treatment was never genuinely attempted, and that gap doesn’t fully close just because the current regimen is stable.
Respecting a competent, clearly stated patient refusal is itself good medicine, not a lapse in guideline adherence — continued pressure would be the actual failure here.
The one thing all three agreed on: this is not a closed question forever — a fall, new cognitive complaint, or other material change would reopen it, and that trigger, not a calendar date, is what was actually written into her chart.