Week 4 case study: insomnia in a 65-year-old woman

The four bullets are a diagnostic reasoning sequence, not the four headings of a SOAP note, and each asks for a different cognitive act. The first asks what history would be pertinent — which means selected, not collected. The second asks what you would examine and order, plus what the case failed to give you, which is an invitation to name a gap rather than praise the workup. The third asks for three differentials with your reasoning and then a final diagnosis with the determination that got you there. The fourth asks for a plan with drug therapy, treatments, education and follow-up. Write them as four arguments; a transcribed SOAP note answers the shape of the assignment without answering any of its questions. Note especially that the third bullet asks two things — the differentials with your reasoning, and then the final diagnosis with how you determined it. That second clause wants the path, not the label.

Dr. Lee hands you one abnormal number before you have asked anything: the hemoglobin A1c has climbed to 8.7%. In a case this tightly written, that is not background. It is the mechanism, offered early, and the history question is really asking whether you will chase it. Poorly controlled diabetes produces osmotic diuresis, osmotic diuresis produces nocturia, and nocturia fragments sleep — which means a patient can present with "I can't sleep" and have a problem that no hypnotic will touch. The same three facts you are given, age 65 plus hypertension plus diabetes, are also the classic risk cluster for obstructive sleep apnoea. Two strong secondary causes are therefore already on the table before you take a single item of history. The wording carries more than the number, too: Dr. Lee says the A1c has climbed. You are being given a trajectory, not a snapshot, and a worsening value alongside a new symptom argues for a causal link more strongly than a static one.

That is the governing principle for the whole case, and it is the one most answers miss: insomnia in an older adult is usually secondary, and primary insomnia disorder is a diagnosis of exclusion. So structure the history to rule causes out rather than to fill in headings. Count the nocturia rather than noting it. Ask what a bed partner has observed, since snoring, witnessed apnoeas and leg kicking are reported by somebody else or not at all. Reconcile the medications with their dosing times, because an antihypertensive regimen that includes an evening diuretic is a fixable cause hiding in plain sight. Then screen mood, pain, reflux, caffeine and alcohol timing, and the actual sleep schedule and its history. Pin down the sleep history itself as well: when it started, what changed around then, how long she lies awake, whether the trouble is falling asleep or staying asleep, and what she now does about it at three in the morning.

Three differentials with your reasoning, and this case supplies more than three good ones. Choose the ones you can argue from the given facts:

The sub-question about what you would have liked and did not get is worth real marks and is usually skipped. Two answers are almost always right in a case like this. The first is a bed-partner or collateral history, because the two conditions with the strongest objective evidence here — apnoea and periodic limb movements — are invisible to the person experiencing them. The second is a two-week sleep diary, which is the core instrument of insomnia assessment and the thing that converts "I can't sleep" into a pattern with a time course. Actigraphy, an Epworth score, ferritin and a TSH are all defensible additions. Name what the gap would change about your reasoning, not just that it was missing. Say what the missing item would have changed. 'A bed-partner history could have moved apnoea from a possibility to the leading diagnosis' shows you know why you wanted it; listing omissions without consequences reads as a checklist.

The plan is where this case is most often failed, and the reason is that the standard answer has been recommended against by name. The American Academy of Sleep Medicine's 2021 guideline carries exactly one strong recommendation: use multicomponent cognitive behavioural therapy for insomnia. The same guideline separately suggests that clinicians not use sleep hygiene as a single-component therapy. So "educate the patient on sleep hygiene" is not a plan of care; it is a component the guideline specifically declines to endorse on its own. CBT-I is the treatment, and in this patient it sits alongside treating the diabetes, because if nocturia is the mechanism then glycaemic control is insomnia treatment. It is also worth saying what CBT-I actually is, since a plan that names it without content is hard to distinguish from one that does not mean it: education about sleep regulation combined with stimulus control instructions, sleep restriction therapy, and cognitive strategies aimed at the beliefs that keep the patient awake.

On drug therapy, two things matter and one of them is very recent. Mrs. Gomez is 65, which places her within the AGS Beers Criteria, the explicit list of medications typically best avoided in older adults — check the current criteria for the hypnotic classes yourself rather than reciting a rule, because that is the reference standard your preceptor will expect you to have opened. And in April 2026 the AASM published a combination-treatment guideline that suggests against CBT-I plus an insomnia medication over CBT-I alone, while suggesting combination over medication alone. Read that carefully: it does not make adding a drug forbidden, but it removes the assumption that adding one improves on CBT-I by itself, and a patient prioritising more total sleep time early is the stated exception. Then close the loop on follow-up: a review interval with the sleep diary in hand gives you something to measure, which is what turns a plan into a plan of care.